Posts for QNX Software Development Platform 6.4.0 Pre-Release Michael Van Reenen http://community.qnx.com/sf/go/post28685 2009-05-05T13:10:17Z 2009-05-05T13:10:17Z

Hi Peter, Did you have to make any changes for graphics to work on the nanoETXpress or did it work 'out-of-the-box? The graphics enumerator only detects devg-poulsbo.so for the integrated graphics - not vesabios - so you would have to manually edit either the enumerator or display.conf to use devg-vesabios.so The 'modeopts=' option in the body of the device entry in display.conf (not photon section) is the active one used by the driver. The 'photon one' is only used to store the active modeopts line if you were to switch between drivers. You could try setting both to the same value and see if it prevents the clearing of the modeopts. Can you try a test: If you boot into textmmode - manually edit display.conf to setup your modeopts line - then run the dispconf utility as follows: dispconf -i 0x8086,0x8108,0 -d poulsbo -g "xres=800,yres=600" Does your display.conf change? (ie modeopts get cleared) We'd like to try to narrow down where it may be getting cleared. Thanks MIke

Michael Van Reenen 2009-05-05T13:10:17Z Peter Weber http://community.qnx.com/sf/go/post28646 2009-05-04T18:15:49Z 2009-05-04T18:15:49Z

Hi Michael, I have the same problem here on an nanoETXexpress using 641 M3. If I started phgrafix, the modeopts are removed. The Board runs fine in 800x480 but I assume there is no h/w accelleration at all. My assumption for this is easy to follow: The Vesa driver is faster than the poulsbo. Other customers with a DigitalLogic board SMA200 are also complaining about this phenomen. BTW: Do I have to use the same modeopts= for both photon section and device section again in display.conf? Could you please explain what this doubled entry of modeopts= is for? Aloha

Peter Weber 2009-05-04T18:15:49Z Michael Van Reenen http://community.qnx.com/sf/go/post27436 2009-04-18T19:59:41Z 2009-04-18T19:59:41Z

Have you tried using the safe mode to not start Photon? This may prevent dispconf from changing your 'modeopts' line in display.conf

Michael Van Reenen 2009-04-18T19:59:41Z soujanya Yelchuri http://community.qnx.com/sf/go/post27044 2009-04-15T12:57:13Z 2009-04-15T12:57:13Z

Hi, I am porting to QNX 6.4.0 for ppc target from QNX 630. I am facing a problem in xml2 library. my program is getting crashed in xmlCleanupCharEncodingHandlers(void)() function. Are there any compatibility issues??

soujanya Yelchuri 2009-04-15T12:57:13Z Warren Deitch http://community.qnx.com/sf/go/post26905 2009-04-14T14:10:42Z 2009-04-14T14:10:42Z

I had the same problem along with others. The fix for PR66114 seems to have fixed it as well as my main problem. See http://community.qnx.com/sf/go/post26754

Warren Deitch 2009-04-14T14:10:42Z soujanya Yelchuri http://community.qnx.com/sf/go/post26890 2009-04-14T13:29:35Z 2009-04-14T13:29:35Z

hi, if i use pidin with mem option its showing random symbols for shared memory objects. If i use the pidin utility of QNX 630 iam getting the proper names pidin -P datasrvr mem pid tid name prio STATE code data stack 1511483 1 datasrvr 10r CONDVAR 556K 536K 8192(516K)* 1511483 2 datasrvr 10r RECEIVE 556K 536K 4096(132K) 1511483 3 datasrvr 10r RECEIVE 556K 536K 4096(132K) 1511483 4 datasrvr 10r RECEIVE 556K 536K 4096(132K) 1511483 5 datasrvr 10r RECEIVE 556K 536K 4096(132K) 1511483 6 datasrvr 10r RECEIVE 556K 536K 12K(132K) 1511483 7 datasrvr 10r RECEIVE 556K 536K 12K(132K) 1511483 8 datasrvr 10r SIGWAITINFO 556K 536K 8192(132K) 1511483 9 datasrvr 10r NANOSLEEP 556K 536K 4096(132K) 1511483 10 datasrvr 10r RECEIVE 556K 536K 4096(132K) 1511483 11 datasrvr 10r RECEIVE 556K 536K 4096(132K) libc.so.3 @fe300000 496K 12K libsocket.so.2 @fe382000 176K 28K libframework.so.1 @fe3b5000 224K 72K libxml2.so.1 @fe3ff000 576K 128K libm.so.2 @fe4af000 164K 12K libCOb.so.1 @fe4db000 760K 16K libcpp.so.4 @fe59d000 428K 32K shmem/SHM_UNITS483 @80100000 ( 0) 4096 shmem/SHM_MISCS483 @80101000 ( 0) 4096 shmem/SHMEM_MUTEX3 @80102000 ( 0) 4096 shmem/SHM_MISCTEX3 @80103000 ( 0) 4096 T_GEN_INFO░H♥·áH♥■ @80104000 ( 0) 92K ODUCT_DETAILS·áH♥■ @8011b000 ( 0) 4096 ITSCT_DETAILS·áH♥■ @8011c000 ( 0) 4096 NK_CONFIG_DATAáH♥■ @8011d000 ( 0) 4096 AILS_INFO_DATAáH♥■ @8011e000 ( 0) 4096 SCLS_INFO_DATAáH♥■ @8011f000 ( 0) 4096 SCLS_INFO_DATAáH♥■ @80120000 ( 0) 4096 # pidin_630 -P datasrvr pid tid name prio STATE Blocked 1511483 1 datasrvr 10r CONDVAR 480cc66c 1511483 2 datasrvr 10r RECEIVE 2 1511483 3 datasrvr 10r RECEIVE 2 1511483 4 datasrvr 10r RECEIVE 2 1511483 5 datasrvr 10r RECEIVE 2 1511483 6 datasrvr 10r RECEIVE 5 1511483 7 datasrvr 10r RECEIVE 5 1511483 8 datasrvr 10r SIGWAITINFO 1511483 9 datasrvr 10r NANOSLEEP 1511483 10 datasrvr 10r RECEIVE 5 1511483 11 datasrvr 10r RECEIVE 2 # pidin_630 -P datasrvr mem pid tid name prio STATE code data stack 1511483 1 datasrvr 10r CONDVAR 556K 536K 8192(516K)* 1511483 2 datasrvr 10r RECEIVE 556K 536K 4096(132K) 1511483 3 datasrvr 10r RECEIVE 556K 536K 4096(132K) 1511483 4 datasrvr 10r RECEIVE 556K 536K 4096(132K) 1511483 5 datasrvr 10r RECEIVE 556K 536K 4096(132K) 1511483 6 datasrvr 10r RECEIVE 556K 536K 12K(132K) 1511483 7 datasrvr 10r RECEIVE 556K 536K 12K(132K) 1511483 8 datasrvr 10r SIGWAITINFO 556K 536K 8192(132K) 1511483 9 datasrvr 10r NANOSLEEP 556K 536K 4096(132K) 1511483 10 datasrvr 10r RECEIVE 556K 536K 4096(132K) 1511483 11 datasrvr 10r RECEIVE 556K 536K 4096(132K) libc.so.3 @fe300000 496K 12K libsocket.so.2 @fe382000 176K 28K libframework.so.1 @fe3b5000 224K 72K libxml2.so.1 @fe3ff000 576K 128K libm.so.2 @fe4af000 164K 12K libCOb.so.1 @fe4db000 760K 16K libcpp.so.4 @fe59d000 428K 32K v/shmem/SHM_UNITS4 @80100000 ( 0) 4096 v/shmem/SHM_MISCS4 @80101000 ( 0) 4096 hmem/SHMEM_MUTEX.2 @80102000 ( 0) 4096 hmem/SHM_MISCTEX.2 @80103000 ( 0) 4096 DUCT_GEN_INFODATA  @80104000 ( 0) 92K KPRODUCT_DETAILSA  @8011b000 ( 0) 4096 _UNITSCT_DETAILSA  @8011c000 ( 0) 4096 _TANK_CONFIG_DATA  @8011d000 ( 0) 4096 DETAILS_INFO_DATA  @8011e000 ( 0) 4096 _MISCLS_INFO_DATA  @8011f000 ( 0) 4096 _MISCLS_INFO_DATA  @80120000 ( 0) 4096 why so? Is there any impact on the system?

soujanya Yelchuri 2009-04-14T13:29:35Z soujanya Yelchuri http://community.qnx.com/sf/go/post26738 2009-04-13T10:08:17Z 2009-04-13T10:08:17Z

i am using ppc related libmudflapth libraries only. still iam getting mf: erroneous reentrancy detected in `__mf_register'

soujanya Yelchuri 2009-04-13T10:08:17Z Nagaraj Salotagi http://community.qnx.com/sf/go/post26702 2009-04-11T00:15:20Z 2009-04-11T00:15:20Z

Hi, I tried modifying the modeopts to the following: "modeopts=/etc/system/config/poulsbo.conf " the above mentioned path is actually where my poulsbo.conf is present. But, after i boot the system and the system crashes the display.conf file gets changed with "modeopts=" I think the dispconf is changing the display.conf. Even when i change the display.conf the result is same. Am i correct in mentioning that dispconf changes the display.conf fiile even if i give the path for poulsbo.conf? Thanks and Regards, Nagaraj.

Nagaraj Salotagi 2009-04-11T00:15:20Z Michael Van Reenen http://community.qnx.com/sf/go/post26698 2009-04-10T21:44:42Z 2009-04-10T21:44:42Z

Taking a closer look at your display.conf: # # Intel Poulsbo # device { photon { driver { drivername=poulsbo modeopts= } } drivername=poulsbo vid=0x8086 did=0x8108 deviceindex=0x0 modeopts= display { xres=1024 yres=768 refresh=60 pixel_format=rgb565 photon { enabled=1 xoffset=0 yoffset=0 cursor=hardware input_group=1 } } It is not actually setup to use your poulsbo.conf. You will need to change 'modeopts=' to specify modeopts to be the fullpath name of your poulsbo.conf file. For example if poulsbo.conf is in /etc/system/config then you would have: modeopts=/etc/system/config/poulsbo.conf I would suggest try this to see if you see any different results.

Michael Van Reenen 2009-04-10T21:44:42Z Nagaraj Salotagi http://community.qnx.com/sf/go/post26671 2009-04-10T12:52:25Z 2009-04-10T12:52:25Z

Hi, I also tried to start photon from my linux machine nut still it hangs the telnet session on the linux machine. The system is crashing while starting the io-graphics server. In the config file there is a setting for DPLL register - how to interpret that setting? does any setting in the conf files can cause this crash? Please help me. Thanks and Regards, Nagaraj.

Nagaraj Salotagi 2009-04-10T12:52:25Z Nagaraj Salotagi http://community.qnx.com/sf/go/post26664 2009-04-10T06:26:52Z 2009-04-10T06:26:52Z

Hi, Thanks for the reply:-) I did tried the telnet but of no use. The following is what i did - please tell me whether whatever i tried is right. 1)I boot up the target in text mode and start inetd for telnet services. 2)I connected to the target using a linux machine using telnet. 3) I was able to login as root from the linux machine. 4)I started photon on the target. 5)After i started the photon by typing "ph" the telnet session on the linux mach hung and i was not able to do anything. Thanks and Regards, Nagaraj.

Nagaraj Salotagi 2009-04-10T06:26:52Z Michael Van Reenen http://community.qnx.com/sf/go/post26619 2009-04-09T20:04:15Z 2009-04-09T20:04:15Z

Do you have another system that you could use to telnet into your target? If so you could boot into safe mode on the target (not starting Photon), then telnet in and start Photon. After this you should be able do a sloginfo to see if there are any messages relating to graphics.

Michael Van Reenen 2009-04-09T20:04:15Z Nagaraj Salotagi http://community.qnx.com/sf/go/post26555 2009-04-09T14:16:59Z 2009-04-09T14:16:59Z

Hi, I haven't called the QX support yet. Thanks and Regards, Nagaraj.

Nagaraj Salotagi 2009-04-09T14:16:59Z Ryan Mansfield http://community.qnx.com/sf/go/post26549 2009-04-09T14:08:30Z 2009-04-09T14:08:30Z

Can you please try the attached ppc libmudflap libraries? Regards, Ryan Mansfield

Ryan Mansfield 2009-04-09T14:08:30Z Elena Laskavaia http://community.qnx.com/sf/go/post26544 2009-04-09T13:59:36Z 2009-04-09T13:59:36Z

Try to ask on general toolchain forum (btw you have 6.4.0 right?) in Core Development Tools http://community.qnx.com/sf/discussion/do/listTopics/projects.toolchain/discussion.core_development_tools soujanya Yelchuri wrote: > I am working on PPC target. > I tried by changing the oder of libc and libmudflap... > but no change .Still getting > > mf: error in dlsym("calloc"): Symbol not found > mf: error in dlsym("calloc"): Symbol not found > > if i remove linking to -lc(its included by default ) > > it says.. > > unknown symbol: _start > unknown symbol: _start > > > Why so? > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post26486 >

Elena Laskavaia 2009-04-09T13:59:36Z Malte Mundt http://community.qnx.com/sf/go/post26534 2009-04-09T13:43:47Z 2009-04-09T13:43:47Z

Have you tried calling your QNX Priority Support Representative? - Malte

Malte Mundt 2009-04-09T13:43:47Z Nagaraj Salotagi http://community.qnx.com/sf/go/post26531 2009-04-09T13:39:55Z 2009-04-09T13:39:55Z

Hi, This is for info. The system i am using has poulsbo display controller and the SDVO interface of poulsbo is used via pipe B. This SDVO output from poulsbo is converted to analog VGA by another chip CH7317A SDVO/RGB DAC fron CHRONTEL.The display interface is VGA. This is the hardware onwhich i am trying to get the poulsbo driver but without success. Please help me as i need to get this working ASAP. Thanks and Regards, Nagaraj.

Nagaraj Salotagi 2009-04-09T13:39:55Z soujanya Yelchuri http://community.qnx.com/sf/go/post26504 2009-04-09T09:29:03Z 2009-04-09T09:29:03Z

i tried to include it to the other module my module is getting crashed. saying mf: erroneous reentrancy detected in `__mf_register' and when i analyse the coredump it says #0 0xfe33eaac in SignalKill () from libc.so.3 #1 0xfe32de98 in raise () from libc.so.3 #2 0xfe32c40c in abort () from libc.so.3 #3 0xfe5a33e4 in begin_recursion_protect1 (pf=0xfe5b53fc "__mf_register") at /home/ryan/spartan/core-dev-tools/tools/gcc/trunk/linux-x86-o-ntoppc/../ ibmudflap/mf-runtime.c:158 #4 0xfe5a5a2c in __mf_register (ptr=0x4803f0d4, sz=4, type=3, name=0x483246f4 "C:/QNX640/target/qnx6/usr/include/cpp/xlocale:315 (use_fac t<std::ctype<char> >) std::_Lockit _Lock") at /home/ryan/spartan/core-dev-tools/tools/gcc/trunk/linux-x86-o-ntoppc/../ ibmudflap/mf-runtime.c:1107 #5 0x48121c74 in std::use_facet<std::ctype<char> > () #6 0xfe3e3428 in std::basic_ios<char, std::char_traits<char> >::init () from libframework.so.1 #7 0xfe6ae22c in std::ios_base::Init::Init () from libcpp.so.4 #8 0xfe6c3660 in __static_initialization_and_destruction_0 () from libcpp.so.4 #9 0xfe6e9808 in __do_global_ctors_aux () from libcpp.so.4 #10 0xfe6a17e8 in _init () from libcpp.so.4 #11 0xfe34078c in init () from libc.so.3 #12 0xfe344314 in ldd () from libc.so.3 #13 0xfe33fbb4 in _start_ () from libc.so.3 Backtrace stopped: previous frame inner to this frame (corrupt stack?) i think its definitely a libc library compatability issue.

soujanya Yelchuri 2009-04-09T09:29:03Z soujanya Yelchuri http://community.qnx.com/sf/go/post26494 2009-04-09T07:37:39Z 2009-04-09T07:37:39Z

in the forum i got to know that there are some changes need to be done for libc (RTLD_NEXT related) to support libmudflap.so.0. Are the changes done??? Can i use libmudflap.so.??? topic related to that http://community.qnx.com/sf/discussion/do/listPosts/projects.toolchain/discussion.core_development_tools.topc1093. Is that the linking issue related to libc. If not done whow to debug memory corruption ???

soujanya Yelchuri 2009-04-09T07:37:39Z soujanya Yelchuri http://community.qnx.com/sf/go/post26486 2009-04-09T05:49:04Z 2009-04-09T05:49:04Z

I am working on PPC target. I tried by changing the oder of libc and libmudflap... but no change .Still getting mf: error in dlsym("calloc"): Symbol not found mf: error in dlsym("calloc"): Symbol not found if i remove linking to -lc(its included by default ) it says.. unknown symbol: _start unknown symbol: _start Why so?

soujanya Yelchuri 2009-04-09T05:49:04Z Nagaraj Salotagi http://community.qnx.com/sf/go/post26410 2009-04-08T15:29:48Z 2009-04-08T15:29:48Z

Hi, Please find attached the graphics, display.conf and poulsbo.conf files which i am using. The system i am working on has poulsbo with external display (VGA coming through a crontel chip) via SDVO-B of the poulsbo chip. I trie 800*600 also which gave me the same result (screen with lot of colors). i am using the GA version of QNX 6.4. Please help me. Thanks and Regards, Nagaraj.

Nagaraj Salotagi 2009-04-08T15:29:48Z Elena Laskavaia http://community.qnx.com/sf/go/post26375 2009-04-08T14:14:18Z 2009-04-08T14:14:18Z

You definetly don't want to include libmalloc.so.2 - it would try redefine same symbols again. What platform you are using? Can you switch order of -lmudflapth and -lc? soujanya Yelchuri wrote: > Hi, > I am trying to debug a memory corruption issue by using libmudflap.so.0 > > I included like this. > > CCFLAGS+=-fmudflapth > LDOPTS+=-lc -fmudflapth -lmudflapth > > > the code is getting compiled properly at the run time iam getting > > mf: error in dlsym("calloc"): Symbol not found > mf: error in dlsym("calloc"): Symbol not found > > I thought i have to include libmalloc.so.2 so i included it... > then its saying > > mf: error in dlsym("map"): Symbol not found > mf: error in dlsym("map"): Symbol not found > > I dont understand why this is comming. > What libraries i have to include with libmudflap.so.0 > can anyone let me know? > > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post26367 >

Elena Laskavaia 2009-04-08T14:14:18Z soujanya Yelchuri http://community.qnx.com/sf/go/post26367 2009-04-08T13:51:42Z 2009-04-08T13:51:42Z

Hi, I am trying to debug a memory corruption issue by using libmudflap.so.0 I included like this. CCFLAGS+=-fmudflapth LDOPTS+=-lc -fmudflapth -lmudflapth the code is getting compiled properly at the run time iam getting mf: error in dlsym("calloc"): Symbol not found mf: error in dlsym("calloc"): Symbol not found I thought i have to include libmalloc.so.2 so i included it... then its saying mf: error in dlsym("map"): Symbol not found mf: error in dlsym("map"): Symbol not found I dont understand why this is comming. What libraries i have to include with libmudflap.so.0 can anyone let me know?

soujanya Yelchuri 2009-04-08T13:51:42Z soujanya Yelchuri http://community.qnx.com/sf/go/post26319 2009-04-08T08:58:15Z 2009-04-08T08:58:15Z

This is the crash message : Process 2117670 (datasrvr) terminated SIGSEGV code=1 fltno=11 ip=fe3c42f0(libframework.so.1@_ZN16CMessageServices14RouteToHandlerEiimjPvRbRi+0x158) mapaddr=0000f2f0. ref=00000000

soujanya Yelchuri 2009-04-08T08:58:15Z Nagaraj Salotagi http://community.qnx.com/sf/go/post26317 2009-04-08T08:44:22Z 2009-04-08T08:44:22Z

Hi, I am having the same problem. Unable to boot into photon. I tried forcing the resolution to 800*600 but of no use. Please help me boot QNX6.4 on a ATOM platform with poulsbo driver. Thanks and Regards, Nagaraj.

Nagaraj Salotagi 2009-04-08T08:44:22Z Nagaraj Salotagi http://community.qnx.com/sf/go/post26316 2009-04-08T08:32:49Z 2009-04-08T08:32:49Z

Hi, I am using a VGA display. I tried QNX6.4 build on the atom but the poulsbo driver seems to be crashing with the screen displaying lot of artifacts. I have put the display.conf and poulsbo.conf from the crownbeach bsp. How to get the slogino output - it so happens that as boot QNX6.4 the system crashes while starting the graphics displaying a screen with lot of artifacts.

Nagaraj Salotagi 2009-04-08T08:32:49Z soujanya Yelchuri http://community.qnx.com/sf/go/post26309 2009-04-08T06:22:16Z 2009-04-08T06:22:16Z

Hi, Thanks for the reply. The crash happens frequently i means one out of 20 to 30 startups. Not every time. And this happens especially in datasrvr module (related to my code) start up. The datasrvr module is getting crashed in 2 different senarios 1)that nano signal issue as i mentioned above. 2)The module is crasing 1 out of 6 to 7 startups at return from a function . When i do the coredump its showing the code part which is not at all related to its start up... The module was working fine on QNX 630 . I dont understand why this is happening. Its look like a memory corruption issue for me. what do u say?

soujanya Yelchuri 2009-04-08T06:22:16Z Douglas Bailey http://community.qnx.com/sf/go/post26286 2009-04-07T20:33:45Z 2009-04-07T20:33:45Z

The "Crash" message is output by the kernel when it detects a situation it never expects to see. Usually it's an indication of a kernel bug. If you're curious, the file and line number point to services/system/ker/nano_signal.c, line 1066, which is in deliver_fault(). There we have the following test: if(pup != NULL) { if((pup->count > 1) || ((pup != first) &&(pup->priority == first->priority) &&(pup->code == first->code) &&(pup->value == first->value) &&(pup->id == first->id))) { // We have double exception - should never happen crash(); } } I think what that code is doing is trying to determine if the kernel detected a fault while it was processing another fault. We have a problem report open which I think may be related -- it reports the crash at line 1081 but the code has evolved since it was opened. Does your crash happen consistently? If so, it could help us track down the root of this problem. Can you tell me what your application is doing at startup when the crash occurs? Doug > -----Original Message----- > From: soujanya Yelchuri [mailto:community-noreply@qnx.com] > Sent: Monday, April 06, 2009 10:09 AM > To: sdp640prerelease-community > Subject: what does this mean ->Crash[0,0] at nano_signal line 1066. > > Hi. > Iam porting to QNX 640 for PPC traget. > At the start up of my application software iam getting > > Crash[0,0] at nano_signal line 1066 and the system is rebooting. > I did not understand what does it mean? > > Can anybody explain? > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post26064 > >

Douglas Bailey 2009-04-07T20:33:45Z soujanya Yelchuri http://community.qnx.com/sf/go/post26064 2009-04-06T14:08:58Z 2009-04-06T14:08:58Z

Hi. Iam porting to QNX 640 for PPC traget. At the start up of my application software iam getting Crash[0,0] at nano_signal line 1066 and the system is rebooting. I did not understand what does it mean? Can anybody explain?

soujanya Yelchuri 2009-04-06T14:08:58Z Michael Van Reenen http://community.qnx.com/sf/go/post26033 2009-04-05T01:21:23Z 2009-04-05T01:21:23Z

Can you post the output of 'sloginfo' after trying to start Photon? What type of display (LVDS, DVI, VGA) are you using? Thanks

Michael Van Reenen 2009-04-05T01:21:23Z Nagaraj Salotagi http://community.qnx.com/sf/go/post25673 2009-03-31T15:04:05Z 2009-03-31T15:04:05Z

> Without BSP, I can run QNX 6.4 in terminal mode, and then if I run ph to enter > photon, ph will failed while it tries to run io-graph. > However, this BSP also provides two config files, display.conf and poulsbo. > conf, as I memtioned in earlier post. > If I run QNX 6.4 without BSP but with this two config files, and with my > modification as I memtioned before, I can enter photon but the result is the > same - overlay. > > As you memtioned above, it seems I still need poulsbo driver to run 3D. > I'm not sure if the problem is the setting of poulsbo.conf. > > general:uselvds=1,usesdvo=0,overlaypipe=0,planecpipe=0,dualview=0 > pipea:pllref=96000,dpll=0x48000c00 > pipeb:pllref=96000,dpll=0x48000c00 > > Do you think if my setting is right? I don't even know how to correctly set > dpll. > > Thanks for you attention.... > Wayne Hi, I have the poulsbo graphic controller vid=0x8086,did=0x8108. Could any body please help me in starting the poulsbo driver given in the Crownbeach BSP. I have QNX6.3.2 installed on which i installed AGTDK2.0.0 and applied the patch ID-331. I have the graphics file in the /etc/system/enum/devices directory with the following entry: device(pci, ven=8086, dev=8108) cfgfile(fname, /etc/system/config/graphics-traplist) echo(io-display "-dvid=0x$(ven),did=0x$(dev)") With this change i am not able to boot into photon the screen displays: login: Cannot attach mouse input report (error code 2) Please help me. Thanks and Regards, Nagaraj

Nagaraj Salotagi 2009-03-31T15:04:05Z Colin Burgess http://community.qnx.com/sf/go/post23207 2009-03-02T15:28:57Z 2009-03-02T15:28:57Z

Which processes are crashing? pidin fds will send _IO_DUP and _IO_FDINFO messages to the resource managers. I'm not sure why pidin channels would have any effect... soujanya Yelchuri wrote: > with 6.4.0 on ppc targets if i use pidin -P fds > or pidin channels > some of the modules are getting crashed. > What may be the reason for this??? > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post22820 > >

Colin Burgess 2009-03-02T15:28:57Z Frederic Plante http://community.qnx.com/sf/go/post22856 2009-02-24T19:51:00Z 2009-02-24T19:51:00Z

Well... given that 6.4.0 was released a few months ago... It seems like this forum is not that much suited anymore... :-) The forum will be made read-only soon, if you have any questions/comments about the SDP 6.4.0, they can be posted to the qnx_momentics_community_support forum or any technology-specific project forums (networking, OS, tools, drivers, etc.). New pre-releases are coming soon (hint 6.4.1) ! :-) Questions about these shall be handled thought a new forum: http://community.qnx.com/sf/discussion/do/listTopics/projects.community/discussion.qnx_software_development_platf_0 Thanks, and enjoy QNX! - Fred

Frederic Plante 2009-02-24T19:51:00Z soujanya Yelchuri http://community.qnx.com/sf/go/post22820 2009-02-24T14:46:58Z 2009-02-24T14:46:58Z

with 6.4.0 on ppc targets if i use pidin -P fds or pidin channels some of the modules are getting crashed. What may be the reason for this???

soujanya Yelchuri 2009-02-24T14:46:58Z soujanya Yelchuri http://community.qnx.com/sf/go/post22818 2009-02-24T14:33:07Z 2009-02-24T14:33:07Z

I am working on ppc target. Thre flow of my code is as below. 1)client request a server. 2)server creates a channel and do an ertry in map of shared memory. 3)client get the server channel info from shred memory and establish the connection. i checked all the three steps everything is fine. example : I found that one of fsm thread is acuring mutex and connectattach is failing when tried to connect to that particular module.(fsm) (pid =428073,channel id = 2,node id =0) here is the info. 4268073 1 fsm 10r CONDVAR (0x4804d7b0) 4268073 2 fsm 10r MUTEX (0x480651a4) 4268073-01 #1 4272170 1 tdm 10r CONDVAR (0x4809c734) 4272170 2 tdm 10r RECEIVE 2 4272170 3 tdm 10r RECEIVE 2 4272170 4 tdm 10r RECEIVE 2 4272170 5 tdm 10r RECEIVE 2 4272170 6 tdm 10r RECEIVE 5 4272170 7 tdm 10r RECEIVE 5 4272170 8 tdm 10r RECEIVE 5 4272170 9 tdm 10r RECEIVE 5 4272170 10 tdm 10r RECEIVE 8 4272170 11 tdm 10r RECEIVE 8 4272170 12 tdm 10r RECEIVE 8 4272170 13 tdm 10r RECEIVE 8 4272170 14 tdm 10r RECEIVE 11 4272170 15 tdm 10r RECEIVE 11 4272170 16 tdm 10r RECEIVE 11 4272170 17 tdm 10r RECEIVE 11 4272170 18 tdm 10r NANOSLEEP 4272170 19 tdm 10r NANOSLEEP 4272170 20 tdm 10r SEM 193add4 4272170 21 tdm 10r NANOSLEEP 4272170 22 tdm 10r NANOSLEEP 4436011 1 bin/pidin 10r REPLY 1 # pidin -P fsm channels pid tid name 4268073 1 fsm 4268073 2 fsm # pidin -P fds pid tid name prio STATE Blocked # pidin -P fsm fds pid name 4268073 fsm 0 4101 rw 0 /dev/ttyp0 1 4101 rw 0 /dev/ttyp0 2 4101 rw 0 /dev/ttyp0 3 4102 rw 0 /fg/FG4TG/logger_queue 0s 1 1s 253981 2s 4227108 the module fsm acquires mutex for long time. when try to connect to it connect attach fails. I did not fine this kind of behaviour in QNX 6.3.0. After migrationg to QNX 6.4.0 some of the modules are acquring mutex for long time.. and program is getting struck @ startup. And the profram getting struck at that particular point. please help me on this.

soujanya Yelchuri 2009-02-24T14:33:07Z Sean Boudreau http://community.qnx.com/sf/go/post22635 2009-02-20T16:33:15Z 2009-02-20T16:33:15Z

On Fri, Feb 20, 2009 at 12:00:49AM -0500, soujanya Yelchuri wrote: > Hi, > My application is working fine with QNX 630. > Recently i am trying to migarte my application to QNX 6.4.0. > My application is is failing sometimes. > When i debuged the issue i found that.. > My function is failing at connectattach() sunction with error no 3. > i.e no channel id or node is or pid...?? > I checked the nid,chid,pid everything is fine.. > Why is it so...??? > Some times its working fine... > Its a stop isse for me... > Help me in solving this.../ It's not really possible to say from the above description. AFAIK there's no known regressions in this area. Do you have a test case? -seanb

Sean Boudreau 2009-02-20T16:33:15Z soujanya Yelchuri http://community.qnx.com/sf/go/post22584 2009-02-20T05:00:44Z 2009-02-20T05:00:44Z

Hi, My application is working fine with QNX 630. Recently i am trying to migarte my application to QNX 6.4.0. My application is is failing sometimes. When i debuged the issue i found that.. My function is failing at connectattach() sunction with error no 3. i.e no channel id or node is or pid...?? I checked the nid,chid,pid everything is fine.. Why is it so...??? Some times its working fine... Its a stop isse for me... Help me in solving this.../

soujanya Yelchuri 2009-02-20T05:00:44Z Sean Boudreau http://community.qnx.com/sf/go/post22524 2009-02-19T14:33:55Z 2009-02-19T14:33:55Z

On Wed, Feb 18, 2009 at 02:45:21PM -0500, hamid marshall wrote: > >> I suspect you're overriding it somewhere. > > Yes, I am overriding it in my startupks.sh to known dir. I mean you must have overridden it to /fs/tmpfs somewhere. Now you're overriding your override. > > Sean, > I have another post related to large file transfers to the imx35 target where the cksum do not match. Have you had a chance to take a look at it. I haven't looked at this one personally. -seanb

Sean Boudreau 2009-02-19T14:33:55Z hamid marshall http://community.qnx.com/sf/go/post22484 2009-02-18T19:45:08Z 2009-02-18T19:45:08Z

>> I suspect you're overriding it somewhere. Yes, I am overriding it in my startupks.sh to known dir. Sean, I have another post related to large file transfers to the imx35 target where the cksum do not match. Have you had a chance to take a look at it. Thanks,

hamid marshall 2009-02-18T19:45:08Z Sean Boudreau http://community.qnx.com/sf/go/post22407 2009-02-18T00:29:28Z 2009-02-18T00:29:28Z

On Tue, Feb 17, 2009 at 05:03:51PM -0500, hamid marshall wrote: > Yap, that did it. The TMPDIR was incorrectly set to /fs/tmpfs where the file system does not exit. I reseted the TMPDIR to a known path and mget is fine now. BTW, where is this TMPDIR path set? It is not in the startks.sh It's set to /tmp in /etc/profile on my system if not already set. I suspect you're overriding it somewhere. -seanb

Sean Boudreau 2009-02-18T00:29:28Z hamid marshall http://community.qnx.com/sf/go/post22406 2009-02-18T00:16:40Z 2009-02-18T00:16:40Z

We have done some "cp -R" of large dirs (3G+) sizes via nfs from a Linux host to a imx35-qnx64 target. We were able to get the complete file transfers (sizes match) to the target board which is imx35 board. However some of these large files on the target, their cksum do not match with the original files in the Linux host. The size of the files that do not match is above 2G. This is probably along the lines of the problem with transferring large files from "Target File System Nav" utility of the Momentics. Thanks,

hamid marshall 2009-02-18T00:16:40Z hamid marshall http://community.qnx.com/sf/go/post22401 2009-02-17T22:03:47Z 2009-02-17T22:03:47Z

Yap, that did it. The TMPDIR was incorrectly set to /fs/tmpfs where the file system does not exit. I reseted the TMPDIR to a known path and mget is fine now. BTW, where is this TMPDIR path set? It is not in the startks.sh

hamid marshall 2009-02-17T22:03:47Z Sean Boudreau http://community.qnx.com/sf/go/post22391 2009-02-17T20:43:15Z 2009-02-17T20:43:15Z

On Tue, Feb 17, 2009 at 02:58:56PM -0500, hamid marshall wrote: > >Is that the exact message? > >Do you have the TMPDIR environmental variable exported > >to '/fs/tmpfs'? > > Hi sean, Yes this is the exact message. I do not have a /fs file system either. The BSP boots to the RAM disk. I start the devb* driver manually and mount my qnx6 partition. I cd to the /mnt/qnx6 where the qnx6 partition is and try the ftp and I get the message above. > > Below is the ls of the dirs at the prompt right after the boot sequence completes, before mounting qnx6 partition > > # ls > bin dev proc tmp usr What about the question WRT TMPDIR? I can reproduct it as follows: /tmp >TMPDIR=/fs/tmpfs ftp 127.1 Connected to 127.1. 220 localhost.localdomain FTP server (QNXNTO-ftpd 20070723) ready. Name (127.1:seanb): 331 Password required for seanb. Password: 230- Welcome to QNX Neutrino! 230 User seanb logged in. Remote system type is UNIX. Using binary mode to transfer files. ftp> mget /home/seanb/t* ftp: unable to create temporary file /fs/tmpfs: No such file or directory ftp>

Sean Boudreau 2009-02-17T20:43:15Z hamid marshall http://community.qnx.com/sf/go/post22390 2009-02-17T20:10:04Z 2009-02-17T20:10:04Z

When I try to do l;arge file (2G+) transfer from the win32 host to qnx6 partition on the target I get "Stream closed" error even the when I have the subnet isolated via switch to the win32 host and the target. Is this expected? Is there a timer value I can set in the File sys Nav to large file transfers. I have been looking at the qnx site, but have not been able to see anything regarding this. Thanks,

hamid marshall 2009-02-17T20:10:04Z hamid marshall http://community.qnx.com/sf/go/post22385 2009-02-17T19:58:52Z 2009-02-17T19:58:52Z

>Is that the exact message? >Do you have the TMPDIR environmental variable exported >to '/fs/tmpfs'? Hi sean, Yes this is the exact message. I do not have a /fs file system either. The BSP boots to the RAM disk. I start the devb* driver manually and mount my qnx6 partition. I cd to the /mnt/qnx6 where the qnx6 partition is and try the ftp and I get the message above. Below is the ls of the dirs at the prompt right after the boot sequence completes, before mounting qnx6 partition # ls bin dev proc tmp usr

hamid marshall 2009-02-17T19:58:52Z soujanya Yelchuri http://community.qnx.com/sf/go/post22338 2009-02-17T13:28:30Z 2009-02-17T13:28:30Z

when i tried to debug the problem some times the connectattach() function is failing. and the error number is ESRCH The node indicated by nd, the process indicated by pid, or the channel indicated by chid doesn't exist. the nid,pid,channel id are fine. Then why its failing? and it does not fail always. some times. Is there any change included in QNX 6.4.0 regarding IPC communication(channel create,connectattach) etc ??? As per release notes no such info. Let me know why its happening.???

soujanya Yelchuri 2009-02-17T13:28:30Z Sean Boudreau http://community.qnx.com/sf/go/post22314 2009-02-16T15:26:02Z 2009-02-16T15:26:02Z

On Sun, Feb 15, 2009 at 04:56:03PM -0500, hamid marshall wrote: > Hi Robert: > > It is a imx35 hardware with QNX6.4 BSP > here is cap from the prompt: > > ftp> bin > 200 Type set to I > ftp> mget * > ftp: unable to create temporary file /fs/tmpfs: No such file or directory > Is that the exact message? Do you have the TMPDIR environmental variable exported to '/fs/tmpfs'? -seanb

Sean Boudreau 2009-02-16T15:26:02Z Robert Craig http://community.qnx.com/sf/go/post22300 2009-02-16T01:09:04Z 2009-02-16T01:09:04Z

Hi Hamid: What is the other endpoint for the FTP session? Are you FTP-ing from the board to a neutrino x86 box or the other way around? What happens if you type in "ls" before the mget? R. -----Original Message----- From: hamid marshall [mailto:community-noreply@qnx.com] Sent: Sun 2/15/2009 4:56 PM To: sdp640prerelease-community Subject: Re: ftp problems on qnx6.4 Hi Robert: It is a imx35 hardware with QNX6.4 BSP here is cap from the prompt: ftp> bin 200 Type set to I ftp> mget * ftp: unable to create temporary file /fs/tmpfs: No such file or directory _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post22299

Robert Craig 2009-02-16T01:09:04Z hamid marshall http://community.qnx.com/sf/go/post22299 2009-02-15T21:56:00Z 2009-02-15T21:56:00Z

Hi Robert: It is a imx35 hardware with QNX6.4 BSP here is cap from the prompt: ftp> bin 200 Type set to I ftp> mget * ftp: unable to create temporary file /fs/tmpfs: No such file or directory

hamid marshall 2009-02-15T21:56:00Z Robert Craig http://community.qnx.com/sf/go/post22298 2009-02-15T17:34:06Z 2009-02-15T17:34:06Z

Under neutrino, tmp is normally under /, so /tmp exists but not /fs/tmp. What platform/CPU are you FTP-ing from or to? Robert ----- Original Message ----- From: hamid marshall <community-noreply@qnx.com> To: sdp640prerelease-community <post22294@community.qnx.com> Sent: Sat Feb 14 23:34:22 2009 Subject: ftp problems on qnx6.4 When I try the mget from the ftp prompt is get the error: "/fs/tmpfs" is not available. My understanding is /fs/tmp is mounted from the BSP/OS image side when the boot sequence completes, similar to NET-BSD. Is this case with qnx6.4 as well? Thanks, _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post22294

Robert Craig 2009-02-15T17:34:06Z hamid marshall http://community.qnx.com/sf/go/post22294 2009-02-15T04:34:18Z 2009-02-15T04:34:18Z

When I try the mget from the ftp prompt is get the error: "/fs/tmpfs" is not available. My understanding is /fs/tmp is mounted from the BSP/OS image side when the boot sequence completes, similar to NET-BSD. Is this case with qnx6.4 as well? Thanks,

hamid marshall 2009-02-15T04:34:18Z Peter Martin http://community.qnx.com/sf/go/post22247 2009-02-13T17:39:33Z 2009-02-13T17:39:33Z

On Fri, 13 Feb 2009, soujanya Yelchuri wrote: >HI, iam using sqlite for my application. When i compiled my application on >QNX 6.4.0 i facing issues with data base. Let me know is there any >compatibility issue of sqlite with QNX 6.4.0 (gcc 4.2.4). > Sqlite is used as part of the Aviage multimedia suite and hasn't shown any specific issues. Can you post more details on the problem and version of sqlute that you are using? Thanks, Peter

Peter Martin 2009-02-13T17:39:33Z Matthias Klein http://community.qnx.com/sf/go/post22181 2009-02-13T09:29:38Z 2009-02-13T09:29:38Z

Which issues do you have ? I made some small tests in the past without problems. (We plan to use SQLite in the future ...) _____________________________________________________________________________________ H. Kleinknecht & Co. GmbH Tel.: +49 (271) 39 83-0 Mail: vertrieb@kleinknecht.de Eiserntalstra?e 358 Fax: +49 (271) 39 83-100 Web: www.kleinknecht.de 57080 Siegen Registergericht / Register Court: Amtsgericht Siegen, HRB 1087 Geschaftsfuhrer / Managing Directors: Rainer Sa?mann, Christoph Hauck _____________________________________________________________________________________ Die vorgenannten Angaben werden jeder E-Mail automatisch hinzugefugt. Dies ist kein Anerkenntnis, dass es sich beim Inhalt dieser E-Mail um eine rechtsverbindliche Erklarung der H. Kleinknecht & Co. GmbH handelt. Erklarungen, die die H. Kleinknecht & Co. GmbH verpflichten, bedurfen jeweils der Unterschrift durch zwei zeichnungsberechtigte Personen.

Matthias Klein 2009-02-13T09:29:38Z soujanya Yelchuri http://community.qnx.com/sf/go/post22179 2009-02-13T09:20:58Z 2009-02-13T09:20:58Z

HI, iam using sqlite for my application. When i compiled my application on QNX 6.4.0 i facing issues with data base. Let me know is there any compatibility issue of sqlite with QNX 6.4.0 (gcc 4.2.4).

soujanya Yelchuri 2009-02-13T09:20:58Z Norton Allen http://community.qnx.com/sf/go/post21852 2009-02-10T02:36:49Z 2009-02-10T02:36:49Z

Go figure, you're right. I had no idea what appropriate anchoring flags would apply there. Anchoring the pane on all four sides to the PtDivider isn't entirely intuitive since we actually want it anchored to a subdivision of the PtDivider. But it works that way, which is what I want. Thanks. Misha Nefedov wrote: > When you drag the sash in a PtDivider, the widgets inside of it gets resized differently (most importantly it "blows-up" the last child) than when a PtDivider is resized by its parent. > > Your "green" container has these anchoring flags: > TOP 2 TOP > BOTTOM 2 TOP* > LEFT 2 LEFT > RIGHT 2 LEFT* > I don't think that the '*' anchors are what you want. Change them to BOTTOM 2 BOTTOM > and RIGHT 2 RIGHT, it should be working "as expected". > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post21798 >

Norton Allen 2009-02-10T02:36:49Z Misha Nefedov http://community.qnx.com/sf/go/post21798 2009-02-09T19:10:46Z 2009-02-09T19:10:46Z

When you drag the sash in a PtDivider, the widgets inside of it gets resized differently (most importantly it "blows-up" the last child) than when a PtDivider is resized by its parent. Your "green" container has these anchoring flags: TOP 2 TOP BOTTOM 2 TOP* LEFT 2 LEFT RIGHT 2 LEFT* I don't think that the '*' anchors are what you want. Change them to BOTTOM 2 BOTTOM and RIGHT 2 RIGHT, it should be working "as expected".

Misha Nefedov 2009-02-09T19:10:46Z Norton Allen http://community.qnx.com/sf/go/post21788 2009-02-09T18:04:22Z 2009-02-09T18:04:22Z

Thanks for the reply. Yes, Pt_AUTO_EXTENT is selected. I just tested this out on 6.3.2 and got the same result. Specifically, the first time you resize the window the PtDivider does not resize until you drag the divider. This is independent of whether or not you have dragged the divider before resizing the window. Once you have resized the window and then dragged the divider, it works as it should and each window resize produces an immediate PtDivider resize. I have attached my application.

Norton Allen 2009-02-09T18:04:22Z Misha Nefedov http://community.qnx.com/sf/go/post21708 2009-02-09T00:29:09Z 2009-02-09T00:29:09Z

What you described seems correct, and I would expect it to work. Do you have the Pt_AUTO_EXTENT (Pt_ARG_CONTAINTER_FLAGS) flag set on your window or the immediate parent of your divider set? Can you send me your sample app to look at?

Misha Nefedov 2009-02-09T00:29:09Z Norton Allen http://community.qnx.com/sf/go/post21704 2009-02-08T01:25:50Z 2009-02-08T01:25:50Z

Forgive me if this isn't the best place to ask. I'm using 6.4.0 and trying to come up to speed on Photon development before beginning a new application. I am trying to familiarize myself with the PtDivider widget. I have a base window which contains a PtDivider that contains two PtPanes. That builds and runs, and I can drag the divider and the two panes resize. Next I have anchored the PtDivider to the window on all four sides and I try resizing the window. The PtDivider does not resize when the window is resized until I drag the divider. As soon as I start the drag, the PtDivider and it's PtPanes resize to fill the window. Is there some option I need to select to have the PtDivider resize when the window resizes? Or is there something I can do on window resize that will cause the PtDivider to resize?

Norton Allen 2009-02-08T01:25:50Z Richard Doucet http://community.qnx.com/sf/go/post21663 2009-02-06T20:07:45Z 2009-02-06T20:07:45Z

AECL - OFFICIAL USE ONLY | À USAGE EXCLUSIF - EACL This fixed my problem! Thanks -----Original Message----- From: Ryan Mansfield [mailto:community-noreply@qnx.com] Sent: February 6, 2009 4:01 PM To: sdp640prerelease-community Subject: Re: Won't run 6.3.2 C++ executables! Richard Doucet wrote: > AECL - OFFICIAL USE ONLY | À USAGE EXCLUSIF - EACL > > Yes Please try the attached libc.so.2. Regards, Ryan Mansfield > > -----Original Message----- > From: Ryan Mansfield [mailto:community-noreply@qnx.com] > Sent: February 6, 2009 3:44 PM > To: sdp640prerelease-community > Subject: Re: Won't run 6.3.2 C++ executables! > > Richard Doucet wrote: >> AECL - OFFICIAL USE ONLY | À USAGE EXCLUSIF - EACL >> >> I guess we are ready sooner the I expected. Is it still possible to get an interim fix to try out? > > Sure, was this for x86? > > Regards, > > Ryan Mansfield > >> -----Original Message----- >> From: Ryan Mansfield [mailto:community-noreply@qnx.com] >> Sent: January 22, 2009 11:55 AM >> To: sdp640prerelease-community >> Subject: Re: Won't run 6.3.2 C++ executables! >> >> Hi Richard, >> >> I believe the problem has been finally tracked down. It should be fixed in the 6.4.0.1 release. Please let me know if you need an interim fix. >> >> Regards, >> >> Ryan Mansfield >> >> _______________________________________________ >> QNX Software Development Platform 6.4.0 Pre-Release >> http://community.qnx.com/sf/go/post20612 >> >> CONFIDENTIAL AND PRIVILEGED INFORMATION NOTICE >> >> This e-mail, and any attachments, may contain information that is >> confidential, subject to copyright, or exempt from disclosure. >> Any unauthorized review, disclosure, retransmission, dissemination or >> other use of or reliance on this information may be unlawful and is >> strictly prohibited. >> >> AVIS D'INFORMATION CONFIDENTIELLE ET PRIVILÉGIÉE >> >> Le présent courriel, et toute pièce jointe, peut contenir de >> l'information qui est confidentielle, régie par les droits d'auteur, >> ou interdite de divulgation. Tout examen, divulgation, >> retransmission, diffusion ou autres utilisations non autorisées de >> l'information ou dépendance non autorisée envers celle-ci peut être >> illégale et est strictement interdite. >> >> >> _______________________________________________ >> QNX Software Development Platform 6.4.0 Pre-Release >> http://community.qnx.com/sf/go/post21651 > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post21652 > CONFIDENTIAL AND PRIVILEGED INFORMATION NOTICE > > This e-mail, and any attachments, may contain information that is > confidential, subject to copyright, or exempt from disclosure. > Any unauthorized review, disclosure, retransmission, dissemination or > other use of or reliance on this information may be unlawful and is > strictly prohibited. > > AVIS D'INFORMATION CONFIDENTIELLE ET PRIVILÉGIÉE > > Le présent courriel, et toute pièce jointe, peut contenir de > l'information qui est confidentielle, régie par les droits d'auteur, > ou interdite de divulgation. Tout examen, divulgation, retransmission, > diffusion ou autres utilisations non autorisées de l'information ou > dépendance non autorisée envers celle-ci peut être illégale et est > strictement interdite. > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post21659 _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post21661 CONFIDENTIAL AND PRIVILEGED INFORMATION NOTICE This e-mail, and any attachments, may contain information that is confidential, subject to copyright, or exempt from disclosure. Any unauthorized review, disclosure, retransmission, dissemination or other use of or reliance on this information may be unlawful and is strictly prohibited. AVIS D'INFORMATION CONFIDENTIELLE ET PRIVILÉGIÉE Le présent courriel, et toute pièce jointe, peut contenir de l'information qui est confidentielle, régie par les droits d'auteur, ou interdite de divulgation. Tout examen, divulgation, retransmission, diffusion ou autres utilisations non autorisées de l'information ou dépendance non autorisée envers celle-ci peut être illégale et est strictement interdite.

Richard Doucet 2009-02-06T20:07:45Z Ryan Mansfield http://community.qnx.com/sf/go/post21661 2009-02-06T20:01:08Z 2009-02-06T20:01:08Z

Richard Doucet wrote: > AECL - OFFICIAL USE ONLY | À USAGE EXCLUSIF - EACL > > Yes Please try the attached libc.so.2. Regards, Ryan Mansfield > > -----Original Message----- > From: Ryan Mansfield [mailto:community-noreply@qnx.com] > Sent: February 6, 2009 3:44 PM > To: sdp640prerelease-community > Subject: Re: Won't run 6.3.2 C++ executables! > > Richard Doucet wrote: >> AECL - OFFICIAL USE ONLY | À USAGE EXCLUSIF - EACL >> >> I guess we are ready sooner the I expected. Is it still possible to get an interim fix to try out? > > Sure, was this for x86? > > Regards, > > Ryan Mansfield > >> -----Original Message----- >> From: Ryan Mansfield [mailto:community-noreply@qnx.com] >> Sent: January 22, 2009 11:55 AM >> To: sdp640prerelease-community >> Subject: Re: Won't run 6.3.2 C++ executables! >> >> Hi Richard, >> >> I believe the problem has been finally tracked down. It should be fixed in the 6.4.0.1 release. Please let me know if you need an interim fix. >> >> Regards, >> >> Ryan Mansfield >> >> _______________________________________________ >> QNX Software Development Platform 6.4.0 Pre-Release >> http://community.qnx.com/sf/go/post20612 >> >> CONFIDENTIAL AND PRIVILEGED INFORMATION NOTICE >> >> This e-mail, and any attachments, may contain information that is >> confidential, subject to copyright, or exempt from disclosure. >> Any unauthorized review, disclosure, retransmission, dissemination or >> other use of or reliance on this information may be unlawful and is >> strictly prohibited. >> >> AVIS D'INFORMATION CONFIDENTIELLE ET PRIVILÉGIÉE >> >> Le présent courriel, et toute pièce jointe, peut contenir de >> l'information qui est confidentielle, régie par les droits d'auteur, >> ou interdite de divulgation. Tout examen, divulgation, retransmission, >> diffusion ou autres utilisations non autorisées de l'information ou >> dépendance non autorisée envers celle-ci peut être illégale et est >> strictement interdite. >> >> >> _______________________________________________ >> QNX Software Development Platform 6.4.0 Pre-Release >> http://community.qnx.com/sf/go/post21651 > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post21652 > CONFIDENTIAL AND PRIVILEGED INFORMATION NOTICE > > This e-mail, and any attachments, may contain information that > is confidential, subject to copyright, or exempt from disclosure. > Any unauthorized review, disclosure, retransmission, > dissemination or other use of or reliance on this information > may be unlawful and is strictly prohibited. > > AVIS D'INFORMATION CONFIDENTIELLE ET PRIVILÉGIÉE > > Le présent courriel, et toute pièce jointe, peut contenir de > l'information qui est confidentielle, régie par les droits > d'auteur, ou interdite de divulgation. Tout examen, > divulgation, retransmission, diffusion ou autres utilisations > non autorisées de l'information ou dépendance non autorisée > envers celle-ci peut être illégale et est strictement interdite. > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post21659

Ryan Mansfield 2009-02-06T20:01:08Z Richard Doucet http://community.qnx.com/sf/go/post21659 2009-02-06T19:56:54Z 2009-02-06T19:56:54Z

AECL - OFFICIAL USE ONLY | À USAGE EXCLUSIF - EACL Yes -----Original Message----- From: Ryan Mansfield [mailto:community-noreply@qnx.com] Sent: February 6, 2009 3:44 PM To: sdp640prerelease-community Subject: Re: Won't run 6.3.2 C++ executables! Richard Doucet wrote: > AECL - OFFICIAL USE ONLY | À USAGE EXCLUSIF - EACL > > I guess we are ready sooner the I expected. Is it still possible to get an interim fix to try out? Sure, was this for x86? Regards, Ryan Mansfield > -----Original Message----- > From: Ryan Mansfield [mailto:community-noreply@qnx.com] > Sent: January 22, 2009 11:55 AM > To: sdp640prerelease-community > Subject: Re: Won't run 6.3.2 C++ executables! > > Hi Richard, > > I believe the problem has been finally tracked down. It should be fixed in the 6.4.0.1 release. Please let me know if you need an interim fix. > > Regards, > > Ryan Mansfield > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post20612 > > CONFIDENTIAL AND PRIVILEGED INFORMATION NOTICE > > This e-mail, and any attachments, may contain information that is > confidential, subject to copyright, or exempt from disclosure. > Any unauthorized review, disclosure, retransmission, dissemination or > other use of or reliance on this information may be unlawful and is > strictly prohibited. > > AVIS D'INFORMATION CONFIDENTIELLE ET PRIVILÉGIÉE > > Le présent courriel, et toute pièce jointe, peut contenir de > l'information qui est confidentielle, régie par les droits d'auteur, > ou interdite de divulgation. Tout examen, divulgation, retransmission, > diffusion ou autres utilisations non autorisées de l'information ou > dépendance non autorisée envers celle-ci peut être illégale et est > strictement interdite. > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post21651 _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post21652 CONFIDENTIAL AND PRIVILEGED INFORMATION NOTICE This e-mail, and any attachments, may contain information that is confidential, subject to copyright, or exempt from disclosure. Any unauthorized review, disclosure, retransmission, dissemination or other use of or reliance on this information may be unlawful and is strictly prohibited. AVIS D'INFORMATION CONFIDENTIELLE ET PRIVILÉGIÉE Le présent courriel, et toute pièce jointe, peut contenir de l'information qui est confidentielle, régie par les droits d'auteur, ou interdite de divulgation. Tout examen, divulgation, retransmission, diffusion ou autres utilisations non autorisées de l'information ou dépendance non autorisée envers celle-ci peut être illégale et est strictement interdite.

Richard Doucet 2009-02-06T19:56:54Z Ryan Mansfield http://community.qnx.com/sf/go/post21652 2009-02-06T19:43:40Z 2009-02-06T19:43:40Z

Richard Doucet wrote: > AECL - OFFICIAL USE ONLY | À USAGE EXCLUSIF - EACL > > I guess we are ready sooner the I expected. Is it still possible to get an interim fix to try out? Sure, was this for x86? Regards, Ryan Mansfield > -----Original Message----- > From: Ryan Mansfield [mailto:community-noreply@qnx.com] > Sent: January 22, 2009 11:55 AM > To: sdp640prerelease-community > Subject: Re: Won't run 6.3.2 C++ executables! > > Hi Richard, > > I believe the problem has been finally tracked down. It should be fixed in the 6.4.0.1 release. Please let me know if you need an interim fix. > > Regards, > > Ryan Mansfield > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post20612 > > CONFIDENTIAL AND PRIVILEGED INFORMATION NOTICE > > This e-mail, and any attachments, may contain information that > is confidential, subject to copyright, or exempt from disclosure. > Any unauthorized review, disclosure, retransmission, > dissemination or other use of or reliance on this information > may be unlawful and is strictly prohibited. > > AVIS D'INFORMATION CONFIDENTIELLE ET PRIVILÉGIÉE > > Le présent courriel, et toute pièce jointe, peut contenir de > l'information qui est confidentielle, régie par les droits > d'auteur, ou interdite de divulgation. Tout examen, > divulgation, retransmission, diffusion ou autres utilisations > non autorisées de l'information ou dépendance non autorisée > envers celle-ci peut être illégale et est strictement interdite. > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post21651

Ryan Mansfield 2009-02-06T19:43:40Z Richard Doucet http://community.qnx.com/sf/go/post21651 2009-02-06T19:42:21Z 2009-02-06T19:42:21Z

AECL - OFFICIAL USE ONLY | À USAGE EXCLUSIF - EACL I guess we are ready sooner the I expected. Is it still possible to get an interim fix to try out? -----Original Message----- From: Ryan Mansfield [mailto:community-noreply@qnx.com] Sent: January 22, 2009 11:55 AM To: sdp640prerelease-community Subject: Re: Won't run 6.3.2 C++ executables! Hi Richard, I believe the problem has been finally tracked down. It should be fixed in the 6.4.0.1 release. Please let me know if you need an interim fix. Regards, Ryan Mansfield _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post20612 CONFIDENTIAL AND PRIVILEGED INFORMATION NOTICE This e-mail, and any attachments, may contain information that is confidential, subject to copyright, or exempt from disclosure. Any unauthorized review, disclosure, retransmission, dissemination or other use of or reliance on this information may be unlawful and is strictly prohibited. AVIS D'INFORMATION CONFIDENTIELLE ET PRIVILÉGIÉE Le présent courriel, et toute pièce jointe, peut contenir de l'information qui est confidentielle, régie par les droits d'auteur, ou interdite de divulgation. Tout examen, divulgation, retransmission, diffusion ou autres utilisations non autorisées de l'information ou dépendance non autorisée envers celle-ci peut être illégale et est strictement interdite.

Richard Doucet 2009-02-06T19:42:21Z Colin Burgess http://community.qnx.com/sf/go/post21602 2009-02-06T15:19:23Z 2009-02-06T15:19:23Z

One of your shared libs appears to have accidentally picked up a reference to a symbol it shouldn't have. How are you building them? You can check for references to the symbol with ntox86-nm or ntox86-readelf -s Colin Jnanesh Kumar wrote: > I am getting following error while compiling my application. I have downloaded evaluation version of QNX SDP6.4.0 which contains the packages below: > - QNX® Neutrino® Realtime Operating System > - QNX® Momentics® Tool Suite > > make[1]: Entering directory `/home/clovis/qnx-atg4.0/asptest/TN320HA/src/app/gmplsnni' > qcc -V4.2.4,gcc_ntox86 -Wl,-lsocket,-lsocket,-L,/opt/clovis//buildtools/i386-pc-nto-qnx6.4.0/local/lib,-L,/opt/clovis//buildtools/i386-pc-nto-qnx6.4.0/lib,-R,/opt/clovis//buildtools/i386-pc-nto-qnx6.4.0/local/lib,-R,/opt/clovis//buildtools/i386-pc-nto-qnx6.4.0/lib,-m,i386nto,--export-dynamic,-rpath-link,/home/clovis/qnx-atg4.0/asptest/target/i386/qnx6.4.0/lib /home/clovis/qnx-atg4.0/asptest/TN320HAA/target/i386/qnx6.4.0/obj/src/app/gmplsnni/clCompAppMain.o -L/home/clovis/qnx-atg4.0/asptest/target/i386/qnx6.4.0/lib -L/home/clovis/qnx-atg4.0/asptest/TN320HA/target/i386/qnx6.4.0/lib -Wl,--start-group -lClTimer -lClBuffer -lClOsal -lClCnt -lClHal -lClCorClient -lClEo -lClIoc -lClRmd -lClUtils -lClAmfClient -lClDbal -lClCkpt -lClEventClient -lClTxnAgent -lClLogClient -lClIdl -lClDebugClient -lClNameClient -lClConfig -lClAmfClient -lClAmsMgmt -lClAmsXdr -Wl,--end-group -L/home/clovis/qnx-atg4.0/clasp/src/ASP/components/ground/client -lClGroundClient -lClGround -lm -L/home/clo vis/qnx-atg4.0/clasp/src/ASP/3rdparty/ezxml/stable -lezxml -o /home/clovis/qnx-atg4.0/asptest/TN320HA/target/i386/qnx6.4.0/bin/gmplsnni > /opt/qnx640/host/linux/x86/usr/bin/ntox86-ld: /home/clovis/qnx-atg4.0/asptest/TN320HA/target/i386/qnx6.4.0/bin/gmplsnni: hidden symbol `__preinit_array_start' in /home/clovis/qnx-atg4.0/asptest/TN320HA/target/i386/qnx6.4.0/bin/gmplsnni is referenced by DSO > /opt/qnx640/host/linux/x86/usr/bin/ntox86-ld: final link failed: Nonrepresentable section on output > cc: /opt/qnx640/host/linux/x86/usr/bin/ntox86-ld error 1 > make[1]: *** [/home/clovis/qnx-atg4.0/asptest/TN320HA/target/i386/qnx6.4.0/bin/gmplsnni] Error 1 > make[1]: Leaving directory `/home/clovis/qnx-atg4.0/asptest/TN320HA/src/app/gmplsnni' > make: *** [gmplsnni] Error 2 > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post21573 -- cburgess@qnx.com

Colin Burgess 2009-02-06T15:19:23Z Jnanesh Kumar http://community.qnx.com/sf/go/post21573 2009-02-06T10:11:53Z 2009-02-06T10:11:53Z

I am getting following error while compiling my application. I have downloaded evaluation version of QNX SDP6.4.0 which contains the packages below: - QNX® Neutrino® Realtime Operating System - QNX® Momentics® Tool Suite make[1]: Entering directory `/home/clovis/qnx-atg4.0/asptest/TN320HA/src/app/gmplsnni' qcc -V4.2.4,gcc_ntox86 -Wl,-lsocket,-lsocket,-L,/opt/clovis//buildtools/i386-pc-nto-qnx6.4.0/local/lib,-L,/opt/clovis//buildtools/i386-pc-nto-qnx6.4.0/lib,-R,/opt/clovis//buildtools/i386-pc-nto-qnx6.4.0/local/lib,-R,/opt/clovis//buildtools/i386-pc-nto-qnx6.4.0/lib,-m,i386nto,--export-dynamic,-rpath-link,/home/clovis/qnx-atg4.0/asptest/target/i386/qnx6.4.0/lib /home/clovis/qnx-atg4.0/asptest/TN320HA/target/i386/qnx6.4.0/obj/src/app/gmplsnni/clCompAppMain.o -L/home/clovis/qnx-atg4.0/asptest/target/i386/qnx6.4.0/lib -L/home/clovis/qnx-atg4.0/asptest/TN320HA/target/i386/qnx6.4.0/lib -Wl,--start-group -lClTimer -lClBuffer -lClOsal -lClCnt -lClHal -lClCorClient -lClEo -lClIoc -lClRmd -lClUtils -lClAmfClient -lClDbal -lClCkpt -lClEventClient -lClTxnAgent -lClLogClient -lClIdl -lClDebugClient -lClNameClient -lClConfig -lClAmfClient -lClAmsMgmt -lClAmsXdr -Wl,--end-group -L/home/clovis/qnx-atg4.0/clasp/src/ASP/components/ground/client -lClGroundClient -lClGround -lm -L/home/clovis/qnx-atg4.0/clasp/src/ASP/3rdparty/ezxml/stable -lezxml -o /home/clovis/qnx-atg4.0/asptest/TN320HA/target/i386/qnx6.4.0/bin/gmplsnni /opt/qnx640/host/linux/x86/usr/bin/ntox86-ld: /home/clovis/qnx-atg4.0/asptest/TN320HA/target/i386/qnx6.4.0/bin/gmplsnni: hidden symbol `__preinit_array_start' in /home/clovis/qnx-atg4.0/asptest/TN320HA/target/i386/qnx6.4.0/bin/gmplsnni is referenced by DSO /opt/qnx640/host/linux/x86/usr/bin/ntox86-ld: final link failed: Nonrepresentable section on output cc: /opt/qnx640/host/linux/x86/usr/bin/ntox86-ld error 1 make[1]: *** [/home/clovis/qnx-atg4.0/asptest/TN320HA/target/i386/qnx6.4.0/bin/gmplsnni] Error 1 make[1]: Leaving directory `/home/clovis/qnx-atg4.0/asptest/TN320HA/src/app/gmplsnni' make: *** [gmplsnni] Error 2

Jnanesh Kumar 2009-02-06T10:11:53Z Ralf Beier http://community.qnx.com/sf/go/post21049 2009-01-30T14:05:13Z 2009-01-30T14:05:13Z

Hi, I'm currently trying to port our QNX632 application to QNX640. This will be done within a Windows XP environment using ant and a modified version of cpptasks supporting qcc. Every once in a while I see following error: [cc] cc: spawn of c:\Dev\QNX640\host\win32\x86\usr\lib\gcc\i386-pc-nto-qnx6.4.0\4.2.4\cc1plus failed: No such file or directory qcc failed with return code 1 Restarting the build process usually solves this failure and compilation resumes. Using the MinGW gcc 3.4.5 compilersuit does not show this error. Is there anything I can do to avoid this error? Thanks, Ralf

Ralf Beier 2009-01-30T14:05:13Z soujanya Yelchuri http://community.qnx.com/sf/go/post21040 2009-01-30T13:18:30Z 2009-01-30T13:18:30Z

is there any change in the implementation of mutex lock??? or is there any change in the start up of procnto module?? i tried with procnto -ml option but no improvements.

soujanya Yelchuri 2009-01-30T13:18:30Z Jason Yee http://community.qnx.com/sf/go/post20974 2009-01-29T10:17:56Z 2009-01-29T10:17:56Z

Hi everyone, would you please give me a specific example for how to use "devc-serusb", inluding use these parameters(vid,did,busno,devno,module....) Thanks very much! Jason

Jason Yee 2009-01-29T10:17:56Z Kushal Koolwal http://community.qnx.com/sf/go/post20950 2009-01-28T18:59:59Z 2009-01-28T18:59:59Z

> I do not believe that it is supported, but you might want to post the output you mean both are not supported - ICH9 and the Ethernet chip? Just doing a community search it seems that there are people who are able to boot on ICH9? http://www.qnx.com/search/index.html?q=ICH9&go.x=0&go.y=0&go=search > of pci -vvv from that box. Then I am sure that somebody can give you a clear > statement. > We are right now in the process of building the system. I am not sure if that box is available. I will try to find that out. Thanks.

Kushal Koolwal 2009-01-28T18:59:59Z Jeevan Mathew http://community.qnx.com/sf/go/post20948 2009-01-28T18:14:16Z 2009-01-28T18:14:16Z

> Hi, > > Sorry for the cross posting at: > http://community.qnx.com/sf/discussion/do/listPosts/projects.community/ > discussion.qnx4_community_support.topc5966 > > After posting I immediately realized that it was the wrong forum. Sorry about > that. > > I am wondering if there is a support for the above two ICH9 (SATA driver) and > Intel's 82574IT (Gigabit Ethernet) chipset > in QNX? > > If yes then which version of QNX supports? > > > Tried searching on http://community.qnx.com/sf/wiki/do/viewPage/projects.bsp/ > wiki/Drivers but could not find any > information. > > Thanks I do not believe that it is supported, but you might want to post the output of pci -vvv from that box. Then I am sure that somebody can give you a clear statement. Jeevan

Jeevan Mathew 2009-01-28T18:14:16Z Jeevan Mathew http://community.qnx.com/sf/go/post20945 2009-01-28T18:11:14Z 2009-01-28T18:11:14Z

> Hi, > I am trying to migrate my application developed on ppc target and QNX 630 to > 640.But some of the modules are wating for mutex for long time at the startup. > This behaviour is not specific to paricular modules. > Its behaving differently for consequent startups. > Why the modules acquring mutex and wating for long time?? > I did not faced any difficulty when i used QNX 630. > What may be the reason. > Iam unable to debug the issue. Just a shot into the dark. You might want to pass -ml option to procnto under QNX640. Better ? Jeevan

Jeevan Mathew 2009-01-28T18:11:14Z soujanya Yelchuri http://community.qnx.com/sf/go/post20898 2009-01-28T06:04:55Z 2009-01-28T06:04:55Z

Hi, I am trying to migrate my application developed on ppc target and QNX 630 to 640.But some of the modules are wating for mutex for long time at the startup.This behaviour is not specific to paricular modules. Its behaving differently for consequent startups. Why the modules acquring mutex and wating for long time?? I did not faced any difficulty when i used QNX 630. What may be the reason. Iam unable to debug the issue.

soujanya Yelchuri 2009-01-28T06:04:55Z Kushal Koolwal http://community.qnx.com/sf/go/post20888 2009-01-27T22:34:35Z 2009-01-27T22:34:35Z

Hi, Sorry for the cross posting at: http://community.qnx.com/sf/discussion/do/listPosts/projects.community/discussion.qnx4_community_support.topc5966 After posting I immediately realized that it was the wrong forum. Sorry about that. I am wondering if there is a support for the above two ICH9 (SATA driver) and Intel's 82574IT (Gigabit Ethernet) chipset in QNX? If yes then which version of QNX supports? Tried searching on http://community.qnx.com/sf/wiki/do/viewPage/projects.bsp/wiki/Drivers but could not find any information. Thanks

Kushal Koolwal 2009-01-27T22:34:35Z Derek Leach http://community.qnx.com/sf/go/post20682 2009-01-23T12:19:22Z 2009-01-23T12:19:22Z

No worries, glad it worked out. -Derek

Derek Leach 2009-01-23T12:19:22Z Dave Florek http://community.qnx.com/sf/go/post20669 2009-01-23T04:19:15Z 2009-01-23T04:19:15Z

Thanks Derek, Setting the display-configuration to "Radeon" from the Photon/pwm desktop (and then picking the correct layer) resolved the issues. Next up: getting mouse and keyboard events propagated into the application. Thanks for your quick response, and sorry to be so long responding! Dave Dave Florek Planet 9 Studios 525 Brannan St, Suite 407 San Francisco, CA 94107 http://www.planet9.com/

Dave Florek 2009-01-23T04:19:15Z Richard Doucet http://community.qnx.com/sf/go/post20616 2009-01-22T15:58:40Z 2009-01-22T15:58:40Z

AECL - OFFICIAL USE ONLY | À USAGE EXCLUSIF - EACL We'll wait for the release; thanks. Richard -----Original Message----- From: Ryan Mansfield [mailto:community-noreply@qnx.com] Sent: January 22, 2009 11:55 AM To: sdp640prerelease-community Subject: Re: Won't run 6.3.2 C++ executables! Hi Richard, I believe the problem has been finally tracked down. It should be fixed in the 6.4.0.1 release. Please let me know if you need an interim fix. Regards, Ryan Mansfield _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post20612 CONFIDENTIAL AND PRIVILEGED INFORMATION NOTICE This e-mail, and any attachments, may contain information that is confidential, subject to copyright, or exempt from disclosure. Any unauthorized review, disclosure, retransmission, dissemination or other use of or reliance on this information may be unlawful and is strictly prohibited. AVIS D'INFORMATION CONFIDENTIELLE ET PRIVILÉGIÉE Le présent courriel, et toute pièce jointe, peut contenir de l'information qui est confidentielle, régie par les droits d'auteur, ou interdite de divulgation. Tout examen, divulgation, retransmission, diffusion ou autres utilisations non autorisées de l'information ou dépendance non autorisée envers celle-ci peut être illégale et est strictement interdite.

Richard Doucet 2009-01-22T15:58:40Z Ryan Mansfield http://community.qnx.com/sf/go/post20612 2009-01-22T15:55:23Z 2009-01-22T15:55:23Z

Hi Richard, I believe the problem has been finally tracked down. It should be fixed in the 6.4.0.1 release. Please let me know if you need an interim fix. Regards, Ryan Mansfield

Ryan Mansfield 2009-01-22T15:55:23Z Michael Van Reenen http://community.qnx.com/sf/go/post20574 2009-01-22T14:07:37Z 2009-01-22T14:07:37Z

Did you change your display.conf to have the pixel_format=rgb565 for vmware? After changing the Windows color depth I believe you may have to re-start VMware (or at least your VMware image) ---------- Mike

Michael Van Reenen 2009-01-22T14:07:37Z Malte Mundt http://community.qnx.com/sf/go/post20562 2009-01-22T09:02:30Z 2009-01-22T09:02:30Z

Hi Mike, I tried to switch from 32 bit to 16 bit color depth on a Windows machine here and QNX 6.4 in VmWare failed to start Photon. - Malte

Malte Mundt 2009-01-22T09:02:30Z Wayne Huang http://community.qnx.com/sf/go/post20554 2009-01-22T02:56:44Z 2009-01-22T02:56:44Z

Without BSP, I can run QNX 6.4 in terminal mode, and then if I run ph to enter photon, ph will failed while it tries to run io-graph. However, this BSP also provides two config files, display.conf and poulsbo.conf, as I memtioned in earlier post. If I run QNX 6.4 without BSP but with this two config files, and with my modification as I memtioned before, I can enter photon but the result is the same - overlay. As you memtioned above, it seems I still need poulsbo driver to run 3D. I'm not sure if the problem is the setting of poulsbo.conf. general:uselvds=1,usesdvo=0,overlaypipe=0,planecpipe=0,dualview=0 pipea:pllref=96000,dpll=0x48000c00 pipeb:pllref=96000,dpll=0x48000c00 Do you think if my setting is right? I don't even know how to correctly set dpll. Thanks for you attention.... Wayne

Wayne Huang 2009-01-22T02:56:44Z Michael Van Reenen http://community.qnx.com/sf/go/post20518 2009-01-21T19:27:28Z 2009-01-21T19:27:28Z

This forum is 'QNX Software Development Platform 6.4.0 Pre-Release' which leads me to believe you may be running an early version of 6.4.0 - opposed to the official release version (GA). Are you able to run QNX 6.4.0 on this board without installing the BSP? There is a devg-poulsbo.so that ships with 6.4.0. Can you try to run Photon with this particular driver to see if the problem still occurs? You're correct in that the egl* apps are 3D demo apps. These will run with devg-vesabios.so using SW 3D rendering but you cannot be running Photon (or any other graphics application) at the same time as there is only 1 display plane/layer. Getting the Poulsbo driver to run would give you access to multiple planes/layers allowing multiple graphics applications to run simultaneously. You would also get HW accelerated 3D rendering - meaning better performance for 3d applications.

Michael Van Reenen 2009-01-21T19:27:28Z Mario Charest http://community.qnx.com/sf/go/post20496 2009-01-21T15:20:54Z 2009-01-21T15:20:54Z

I've never had to fiddle with config files to get QNX to work in vmware > -----Original Message----- > From: Michael Van Reenen [mailto:community-noreply@qnx.com] > Sent: January-21-09 10:14 AM > To: sdp640prerelease-community > Subject: Re: installing 6.4.0 in vmware workstation 6.5 > > What color depth is your host running at? QNX 6.4.0 has the VMware > driver default to 32-bit (ARGB8888) with the assumption that Windows is > running in 32-bit. > > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post20494 >

Mario Charest 2009-01-21T15:20:54Z Michael Van Reenen http://community.qnx.com/sf/go/post20494 2009-01-21T15:13:28Z 2009-01-21T15:13:28Z

What color depth is your host running at? QNX 6.4.0 has the VMware driver default to 32-bit (ARGB8888) with the assumption that Windows is running in 32-bit.

Michael Van Reenen 2009-01-21T15:13:28Z Wayne Huang http://community.qnx.com/sf/go/post20473 2009-01-21T12:36:55Z 2009-01-21T12:36:55Z

Sorry, I don't really understand what you mean by GA version of 6.4 or a pre-release version. I installed QNX 6.4 into my machine. The bsp is bsp-intel_crowbeach_pub-1.0.0-20080618-binsrc.zip, which can be downloaded from http://community.qnx.com/sf/frs/do/listReleases/projects.bsp/frs.intel_crown_beach_crb This BSP provides two patches, Core Advanced Graphics TDK Patch (Patch ID 331) and Image-GLES-AGTDK Patch (Patch ID 1048) Experimental. I didn't download them because I don't have Advanced Graphics TDK so that I can't compile them. I unziped bsp-intel_crowbeach_pub-1.0.0-20080618-binsrc.zip, and then you can a prebuilt folder. IThe devg-poulsbo.so I'm using is in the prebuilt folder. Yes, I can the entire display because I can the bottom bar, Launch mene and the clock. If I don't expand the side-bar, it seems stable. Generally, the background of the side-bar is gray; however, it is transparent. I've tried to use 800x600, but ph would fail. I can't even enter photon. About 1024x600, I of course have tried it, but the result is the same as 768. And the other thing is I've tried to use vesabios with 800x600 and 1024x768. It worked well. The background of the side-bar is gray, and they don't overlay. But, if I use vesabios, I will have a problem with 3D. I hope I'm not wrong that egl-* commands are 3D test tools. I run them within vesabios driver, all of them failed... That is why I have to use pousblo driver; otherwise, vesabios would be ok to me.... Thanks for your replying..... Wayne

Wayne Huang 2009-01-21T12:36:55Z Norton Allen http://community.qnx.com/sf/go/post20411 2009-01-20T16:42:20Z 2009-01-20T16:42:20Z

To answer my own question, I found the hint I needed in the VMWare project support forum. I needed to edit display.conf and change the pixel format line in the vmware stanza to read 'pixel_format=rgb565'. Now I appear to be good to go with a vanilla installation of 6.4.0 on vmware workstation 6.5.

Norton Allen 2009-01-20T16:42:20Z Norton Allen http://community.qnx.com/sf/go/post20287 2009-01-18T22:21:14Z 2009-01-18T22:21:14Z

That is "vmware workstation 6.5"

Norton Allen 2009-01-18T22:21:14Z Norton Allen http://community.qnx.com/sf/go/post20286 2009-01-18T22:20:29Z 2009-01-18T22:20:29Z

I just installed vmware 6.5 with QNX 6.4.0, and all I get for photon is "unable to start graphics driver(s)" I editted /etc/system/config/display.conf to comment out all but the vmware driver to no effect. Any other things I should be trying? Has anyone gotten this combination to work?

Norton Allen 2009-01-18T22:20:29Z Michael Van Reenen http://community.qnx.com/sf/go/post20235 2009-01-16T17:50:40Z 2009-01-16T17:50:40Z

Are you using the GA version of 6.4 or a pre-release version? If you don't expand the side-bar does the display otherwise seem stable? Do you see the entire display - your configuration file shows you running 1024x768 while you mention the maximum resolution for the display is 1024x600? Are you able to run 800x600? If so do you see the same problem? Thanks

Michael Van Reenen 2009-01-16T17:50:40Z Derek Leach http://community.qnx.com/sf/go/post20203 2009-01-16T14:13:36Z 2009-01-16T14:13:36Z

I think you are still using the svga driver ... because I believe any Radeon has more than 1 layer. fyi, you can run egl-gears like this: egl-gears -layer=# (where # is a different layer id) btw, io-graphics is the process which attaches to the layer, not the window manager (just for your notes). Run Photon, then right click on the desktop, select Configure, then Display. See what is says there, svga, or radeon. If it says svga, change it to Radeon, then retry your apps. If you still have issues, report back here with the following: 1) your entire /etc/system/config/display.conf 2) output from: pidin mem 3) output from: sloginfo (after you tried to run your application unsuccessfully, if that happens). -Derek

Derek Leach 2009-01-16T14:13:36Z Dave Florek http://community.qnx.com/sf/go/post20182 2009-01-16T01:00:14Z 2009-01-16T01:00:14Z

I'm also running into problems combining OpenGLES and GF, while running Photon and pwm. Specifically, when running egl-gears (as root), I get: # egl-gears gf_layer_attach() failed Writing and debugging my own app, I determined that: (1) My display-device (a Radeon 9700 Pro) has only one display configured, the display has only one layer, and the layer is already attached (presumably by the window-manager). (2) I can attach to the layer using flags ( GF_LAYER_ATTACH_PASSIVE | GF_LAYER_ATTACH_NODEFAULTS ), and proceed according to the various how-to documents, but eventually hit a roadblock with (and forgive the extra output, but in case it helps...): # ./myprog DisplayDevice: number of displays: 1 display 0: 1024X768, 60 Hz number of layers: 1 main layer: 0 gf_layer_attach() failed: layer already in use, re-trying with PASSIVE flag.... width=640, height=480, format=0x00001710 bits per pixel: 16 packing: interleaved unknown-endian has alpha: yes color-space: direct rgb gf_3d_target_create() failed: out of memory: No such process ... the last line referring to the returned code GF_ERR_MEM, and then the output of perror() is "No such process". In addition, it appears that the Photon start-up process at boot-time is rewriting /etc/system/config/display.conf -- I added a block at the top of the file something like: device { drivername=radeon vid=0x1002 did=0x4e44 #Note, 'pci' lists a "secondary (Other)" display at did=0x4e64 deviceindex=0 display { xres=1280 yres=1024 refresh=60 pixel_format=argb1555 photon { enabled=1 xoffset=0 yoffset=0 cursor=hardware input_group=1 } } } and after a reboot, it had been encapsulated to read something like: device { photon { driver { drivername=svga modeopts= } driver { drivername=radeon modeopts= } driver { drivername=vesabios modeopts= } } drivername=svga vid=0x1002 did=0x4e44 #Note, 'pci' lists a "secondary (Other)" display at did=0x4e64 deviceindex=0x0 modeopts= display { xres=1280 yres=1024 refresh=60 pixel_format=argb1555 photon { enabled=1 xoffset=0 yoffset=0 cursor=hardware input_group=1 } } } Finally, I downloaded the sample linked elsewhere in this thread, built it, but get a different runtime error: # ./gf-ph-3d_g Raw Widget: width = 241, height = 172 PdCreateOffscreenContextGF() failed I hope any or all of this makes it obvious what I need to do next! Thanks in advance, Dave Dave Florek Planet 9 Studios 525 Brannan St, Suite 407 San Francisco, CA 94107 415-348-1200 http://www.planet9.com/

Dave Florek 2009-01-16T01:00:14Z Wayne Huang http://community.qnx.com/sf/go/post20089 2009-01-15T03:05:49Z 2009-01-15T03:05:49Z

The plateform I'm using is menlow, Poulsbo SCH. The screen size and the machine shape are very similiar to the game machine, PSP. Therefore, the board connects to the sreen through by LVDS but not DVI or TFT. And, the maximum of the resolution is 1024x600. After I failed with the installation of the current poulsbo graphic driver, I found Bspdown_crownbeach including another pousblo graphic driver. http://community.qnx.com/sf/wiki/do/viewPage/projects.bsp/wiki/Bspdown_crownbeach This BSP graphic driver offers a configuration file, poulsbo.conf. My display.conf is showed as below. device { drivername=poulsbo vid=0x8086 did=0x8108 deviceindex=0 modeopts=/etc/system/config/poulsbo.conf display { xres=1024 yres=768 refresh=60 pixel_format=argb8888 } } The original poulsbo file is: general:uselvds=0,usesdvo=1,overlaypipe=0,planecpipe=0,dualview=0 pipea:pllref=96000,dpll=0x44000c00 pipeb:pllref=96000,dpll=0x48000c00 Because this machine doesn't have SDVO but only has LDVS, I changed the original poulsbo.conf to this general:uselvds=1,usesdvo=0,overlaypipe=0,planecpipe=0,dualview=0 pipea:pllref=96000,dpll=0x48000c00 pipeb:pllref=96000,dpll=0x48000c00 With these display.conf and poulsbo.conf, the BSP graphic driver finally works. I can enter Photon server and see the desktop. However, another problem appears. The menu bar on the right hand side of the desktop, such as internet, Utilities, Games and Configure, doesn't work well. When you click Utilities, it will expand its sub-list, and Games will shift down. Utilities does expand its sub-list; however, Games doesn't shift down. As a result, Utilities' sub-list overlay with Games. If you execute a program like Video Poker in Games, you can only see the window frame of it. If you move it, it won't re-draw. If the original poulsbo graphic driver works with poulsbo.conf, the result is the same. I suspect if the setting of dpll is not proper, but I don't know how to set it. At least I couldn't find any information or discussion telling me how to set it. Thanks for your replying... Wayne

Wayne Huang 2009-01-15T03:05:49Z Nikolsky Oleg http://community.qnx.com/sf/go/post19711 2009-01-09T07:42:49Z 2009-01-09T07:42:49Z

Thank You very much, Derek Leach!

Nikolsky Oleg 2009-01-09T07:42:49Z Derek Leach http://community.qnx.com/sf/go/post19686 2009-01-08T20:12:39Z 2009-01-08T20:12:39Z

> Thank You for reply. > > Can You explain me how can I get fixed version of QNX 6.4.* image like placed > at http://www.qnx.com/download/ for downloading ? Is this downlodable image > restored as new fixies are released? > > O.Nikolsky I _believe_ this is where new releases will be posted, but I am not part of marketing or sales ... I think the best advice I can give is to contact your support representative and give the PR #'s to them. I think if you are entitled to priority support patches(PSP - priority customers), you can request a PSP, otherwise the fixes will be available during the next release cycles.

Derek Leach 2009-01-08T20:12:39Z Nikolsky Oleg http://community.qnx.com/sf/go/post19663 2009-01-08T17:00:08Z 2009-01-08T17:00:08Z

Thank You for reply. Can You explain me how can I get fixed version of QNX 6.4.* image like placed at http://www.qnx.com/download/ for downloading ? Is this downlodable image restored as new fixies are released? O.Nikolsky

Nikolsky Oleg 2009-01-08T17:00:08Z Derek Leach http://community.qnx.com/sf/go/post19629 2009-01-08T14:14:13Z 2009-01-08T14:14:13Z

PR 64417 filed for 6.4.1 PR 64418 filed for 6.4.0.1 Filed as regression.

Derek Leach 2009-01-08T14:14:13Z Derek Leach http://community.qnx.com/sf/go/post19624 2009-01-08T13:50:51Z 2009-01-08T13:50:51Z

yikes! I'll take a quick look, thanks.

Derek Leach 2009-01-08T13:50:51Z Michael Van Reenen http://community.qnx.com/sf/go/post19432 2009-01-06T13:54:12Z 2009-01-06T13:54:12Z

What specific board/platform are you using? What type of display do you have (LCD, TFT, DVI)? Thanks

Michael Van Reenen 2009-01-06T13:54:12Z Nikolsky Oleg http://community.qnx.com/sf/go/post19340 2009-01-05T13:00:54Z 2009-01-05T13:00:54Z

Simple program generates fonts stem names. Its results differ for QNX 6.3.2 and 6.4.0. ------------------------- The program itself: #include <Pt.h> #define nFONTLIST_SIZE 100 int main() { FontName cur_fnt; FontDetails tsFontList [nFONTLIST_SIZE]; int all_fonts, sCurrFont, fonts; PtInit(NULL); if (fonts = PfQueryFonts(PHFONT_ALL_SYMBOLS, PHFONT_ALL_FONTS, tsFontList, nFONTLIST_SIZE) == -1) { perror ("PfQueryFonts() failed: "); PtExit(-1); } sCurrFont=0, all_fonts=0; while(strlen(tsFontList[sCurrFont].desc)) { printf("*%d %s - ", all_fonts, tsFontList[sCurrFont].desc); if (PfGenerateFontName( tsFontList[sCurrFont].desc, 0, 14, cur_fnt) == NULL) { printf("\r-?? %s PfGenerateFontName() failed ?????????????????\n", tsFontList[sCurrFont].desc); } else { printf("%s\n", cur_fnt); all_fonts++; } sCurrFont++; } printf("%d fonts available\n", all_fonts); PtExit(0); } ------------------------- All results for QNX 6.3.2 are correct. Strange mismatching results for QNX 6.4.0 (marked with ???): *0 arial - arial14 *1 BalloonFont - BalloonFont14 *2 Bitstream SB Gothic HK Uni - bitstreamsbgothichkunis14 . . . *29 PC Sanserif - pcss1614 ??? *30 PC Serif - pcs1214 ??? *31 PC Terminal - pcterm1214 ??? *32 Photon Cursor - phcursor14 . . . *41 Symbol - symb2414 ??? *42 Symbol Set BT - symbolsetbts14 *43 Technical - tech1414 ??? -?? term PfGenerateFontName() failed ????????????????? *44 TextFont - TextFont14 . . . *50 Web - webs14 *51 web - webs14 52 fonts available ------------------------- There is also little inaccuracy in HELP on “Fonts” item – FontDetails structure described in file <font_types.h> not in <photon/Pf.h>

Nikolsky Oleg 2009-01-05T13:00:54Z Wayne Huang http://community.qnx.com/sf/go/post19273 2008-12-29T10:04:37Z 2008-12-29T10:04:37Z

Hi all, My graphic card is Poulsbo. While booting up, the system just dies.The screen turns not only white or black but with many colors.I've looked into two files... 1. /etc/system/enum/devices/graphicsThis file includes the following lines. I don't see any problems.device(pci, ven=8086, dev=8108)start/wait( dispconf -i "0x$(ven),0x$(dev),$(index)" -d poulsbo -g "xres=800,yres=600" )requires/wait( $(IODISPLAY_CMD), )mount(-Tio-display "vid=0x$(ven),did=0x$(dev),deviceindex=$(index)" ) 2. /usr/bin/phIn this script file, while running io-graphics, the system dies. I can't find any discussion on internet. No one talks this graphic card.Please tell me if you have any suggestions. Thanks.... Wayne _________________________________________________________________ Show them the way! Add maps and directions to your party invites. http://www.microsoft.com/windows/windowslive/events.aspx

Wayne Huang 2008-12-29T10:04:37Z soujanya Yelchuri http://community.qnx.com/sf/go/post19227 2008-12-24T15:09:02Z 2008-12-24T15:09:02Z

Hi, I am working on PPC platform and using QNX 640. I am facing issues in qnet mounting on en0 with io-pkt. Here is my startup procedure io-pkt-v4 ethmaccmd="mount -T io-pkt -o channel=3,mac=$ethmac devn-ppc8260-nxa.so" $ethmaccmd #waitfor /dev/io-net/en0 #sleep 2 echo "." tv_waitfor -p /dev/io-net/en0 -t 5 -e 1 -v ethmaccmd="mount -T io-pkt -o channel=1,mac=$syncportmac devn-ppc8260-nxa.so" $ethmaccmd #waitfor /dev/io-net/en1 #sleep 2 echo "." tv_waitfor -p /dev/io-net/en1 -t 5 -e 1 -v ethmaccmd="mount -T io-pkt -o channel=2,mac=$svcportmac devn-ppc8260-nxa.so" $ethmaccmd #waitfor /dev/io-net/en2 echo "."tv_waitfor -p /dev/io-net/en2 -t 5 -e 1 -v echo "." mount -T io-pkt -o bind=en0 lsm-qnet.so my driver info is fine i checked with nicinfo. When i mount QNET iam able to see all the other kits under /net folder. When i do pidin -n (other kit ) pidin -n /net/241-NXA820/ pidin -n /net/213-NXA820/ iam getting the following slog information # sloginfo Time Sev Major Minor Args Mar 04 23:04:00 7 15 0 qnet(kif): server_lookup(): invalid scoid 41, 0 Mar 04 23:04:00 7 15 0 qnet(QOS): tx_xmit_init_conn_pkt(): to nd 1 on L4 0 Mar 04 23:04:00 7 15 0 qnet(kif): server_lookup(): invalid scoid 42, 4 2 Mar 04 23:04:46 7 15 0 qnet(L4): lr_resolve(): starting for nd 3 219* .pcm.endress.com Mar 04 23:04:46 7 15 0 qnet(L4): l4_resolve_periodic_en_ionet(): timeo ut: re-tx 1 looking for nd 3 219*.pcm.endress.com Mar 04 23:04:47 7 15 0 qnet(L4): l4_resolve_periodic_en_ionet(): timeo ut: re-tx 2 looking for nd 3 219*.pcm.endress.com Mar 04 23:04:47 7 15 0 qnet(L4): l4_resolve_periodic_en_ionet(): 2 ret ries exceeded for nd 3 219*.pcm.endress.com found 0 interfaces Mar 04 23:04:47 7 15 0 qnet(L4): lr_resolve(): starting for nd 4 219* .pcm.endress.com Mar 04 23:04:47 7 15 0 qnet(L4): l4_resolve_periodic_en_ionet(): timeo ut: re-tx 1 looking for nd 4 219*.pcm.endress.com Mar 04 23:04:47 7 15 0 qnet(L4): l4_resolve_periodic_en_ionet(): timeo ut: re-tx 2 looking for nd 4 219*.pcm.endress.com Mar 04 23:04:47 7 15 0 qnet(L4): l4_resolve_periodic_en_ionet(): 2 ret ries exceeded for nd 4 219*.pcm.endress.com found 0 interfaces Mar 04 23:04:47 7 15 0 qnet(L4): lr_resolve(): starting for nd 5 219* .pcm.endress.com Mar 04 23:04:48 7 15 0 qnet(L4): l4_resolve_periodic_en_ionet(): timeo ut: re-tx 1 looking for nd 5 219*.pcm.endress.com Mar 04 23:04:48 7 15 0 qnet(L4): l4_resolve_periodic_en_ionet(): timeo ut: re-tx 2 looking for nd 5 219*.pcm.endress.com Mar 04 23:04:48 7 15 0 qnet(L4): l4_resolve_periodic_en_ionet(): 2 ret ries exceeded for nd 5 219*.pcm.endress.com found 0 interfaces Mar 04 23:04:51 7 15 0 qnet(QOS): tx_xmit_init_conn_pkt(): to nd 2 on L4 0 Mar 04 23:04:52 7 15 0 qnet(QOS): tx_xmit_init_conn_pkt(): to nd 2 on L4 0 retry 1 Mar 04 23:04:52 7 15 0 qnet(QOS): tx_xmit_init_conn_pkt(): to nd 2 on L4 0 retry 2 Mar 04 23:04:53 7 15 0 qnet(QOS): tx_xmit_init_conn_pkt(): no more L4s to try for nd 2 Mar 04 23:04:53 7 15 0 qnet(QOS): tx_complete(): callback for nd 2 to layer 0 with errno 264 Mar 04 23:04:53 7 15 0 qnet(kif): kif_server_outbound_failed(): NTONET _CONNECT server busy 480e8180 Mar 04 23:04:53 7 15 0 qnet(QOS): nd_change_notify(): Node Down: nd 2 249-NXA820.pcm.endress.com why it is saying node down ?? and server busy??

soujanya Yelchuri 2008-12-24T15:09:02Z Raja chandru http://community.qnx.com/sf/go/post19212 2008-12-24T06:33:28Z 2008-12-24T06:33:28Z

Hello All, I have fixed all the issues in running regression test on sqlite3 test environment. Steps to run the regression test (*) Download the sqlite3 source code from www.sqlite.org (*) Copy the zip file into target (*) run the following commands one by one tar xzf sqlite.tar.gz ;# Unpack the source tree into "sqlite" mkdir bld ;# Build will occur in a sibling directory cd bld ;# Change to the build directory ../sqlite/configure ;# Run the configure script make fulltest ;# Run the makefile. make install ;# (Optional) Install the build products (*) This will create a executive file 'testfixture' in the bld folder (*) Copy the 'testfixture' into /usr/local/bin (*) Then navigate to the path where the sqlite3 test files are located (ex : /home/sqlite3.x.x/test/) (*) Run the test by using # testfixture all.test command . (*) This will run the entire test scripts (*) If u want u can redirect the result to a text file or log file. Thanks and regards Chandru.K

Raja chandru 2008-12-24T06:33:28Z Dave Bott(deleted) http://community.qnx.com/sf/go/post19190 2008-12-23T17:39:27Z 2008-12-23T17:39:27Z

I see that devn-e1000 is checked in on the devn i.e. io-net tree. I have a customer that wants support for the i82575 NIC - the e1000 driver has this, our i82544 driver does not. What is this code doing in the tree if we can't use it for customers ? Can I give a binary out ? Thanks Dave

Dave Bott(deleted) 2008-12-23T17:39:27Z Sean Boudreau http://community.qnx.com/sf/go/post19129 2008-12-22T13:10:49Z 2008-12-22T13:10:49Z

> I am using a QNX 640 pre - release Version. > Let me know the Official release Date for QNX 640. It was released on Nov 3. -seanb

Sean Boudreau 2008-12-22T13:10:49Z soujanya Yelchuri http://community.qnx.com/sf/go/post19126 2008-12-22T12:27:19Z 2008-12-22T12:27:19Z

I am using a QNX 640 pre - release Version. Let me know the Official release Date for QNX 640. Regards, Soujanya Yelchuri Software Engineer | Tools & Embedded | KPIT Cummins | +91 20 66525000 (Ex:2578) | +91 09623441395 (Mob) | Soujanya.yelchuri@Kpitcummins.com <mailto:Soujanya.yelchuri@Kpitcummins.com> | www.kpitcummins.com <http://www.kpitcummins.com/> This message contains information that may be privileged or confidential and is the property of KPIT Cummins. It is intended only for the person to whom it is addressed. If you are not the intended recipient, you are not authorized to read, print, retain copy, disseminate, distribute, or use this message or any part thereof. If you receive this message in error, please notify the sender immediately and delete all copies of this message. KPIT Cummins does not accept any liability for virus infected mails

soujanya Yelchuri 2008-12-22T12:27:19Z Colin Burgess http://community.qnx.com/sf/go/post18934 2008-12-18T14:41:23Z 2008-12-18T14:41:23Z

It would not help with the dependency check failure. But it's the way to add options around libs soujanya Yelchuri wrote: > I tried that option but did not help...! > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post18911 > -- cburgess@qnx.com

Colin Burgess 2008-12-18T14:41:23Z soujanya Yelchuri http://community.qnx.com/sf/go/post18911 2008-12-18T03:56:56Z 2008-12-18T03:56:56Z

I tried that option but did not help...!

soujanya Yelchuri 2008-12-18T03:56:56Z Colin Burgess http://community.qnx.com/sf/go/post18820 2008-12-17T14:44:31Z 2008-12-17T14:44:31Z

BTW - if you want to wrap a particular lib with options, you can use LIBPREF_<libname> and LIBPOST_<libname> eg LIBS = a b c LIBPREF_b = -Bstatic LIBPOST_b = -Bdynamic will add -la -Bstatic -lb -Bdynamic -lc to the link line soujanya Yelchuri wrote: > Thanks a lot. > The issue is solved. > > The problem is solved by giving LDOPTS+= -Bstatic -llibrary name. > > But i dont understand why i should add it externally... > > I did not face this issue when i used 632. > > Regards, > Soujanya. > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post18795 > -- cburgess@qnx.com

Colin Burgess 2008-12-17T14:44:31Z soujanya Yelchuri http://community.qnx.com/sf/go/post18795 2008-12-17T07:15:30Z 2008-12-17T07:15:30Z

Thanks a lot. The issue is solved. The problem is solved by giving LDOPTS+= -Bstatic -llibrary name. But i dont understand why i should add it externally... I did not face this issue when i used 632. Regards, Soujanya.

soujanya Yelchuri 2008-12-17T07:15:30Z Aaron Cripps http://community.qnx.com/sf/go/post18697 2008-12-16T14:27:36Z 2008-12-16T14:27:36Z

My bad ... thanks for pointing that out Ryan. -Aaron On Tuesday 16 December 2008 09:21:12 Ryan Mansfield wrote: > Aaron Cripps wrote: > > sorry, I may have missed one thing earlier ... if you have any extra > > options, you will want to add them as a flag under CCOPTS > > so,: > > LIBS += socket > > CCOPTS += -Bdynamic > > will use the -B flag on the command line, and include socket as a lib > > against which to link. > > -Aaron > > Linker options should be specified with LDOPTS. > > > On Tuesday 16 December 2008 00:54:20 soujanya Yelchuri wrote: > >> If a libarary which is included without any option in LIBS+= path in > >> common.mk the compiler is searching for its .a. But by default it > >> should search for .so in the included paths. > > If you add a library to LIBS, it is the makefiles that search for the .a > to do dependency checking. By default the linker will against the shared > library. > > Regards, > > Ryan Mansfield > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post18695

Aaron Cripps 2008-12-16T14:27:36Z Ryan Mansfield http://community.qnx.com/sf/go/post18695 2008-12-16T14:21:10Z 2008-12-16T14:21:10Z

Aaron Cripps wrote: > sorry, I may have missed one thing earlier ... if you have any extra options, > you will want to add them as a flag under CCOPTS > so,: > LIBS += socket > CCOPTS += -Bdynamic > will use the -B flag on the command line, and include socket as a lib against > which to link. > -Aaron Linker options should be specified with LDOPTS. > On Tuesday 16 December 2008 00:54:20 soujanya Yelchuri wrote: >> If a libarary which is included without any option in LIBS+= path in >> common.mk the compiler is searching for its .a. But by default it should >> search for .so in the included paths. If you add a library to LIBS, it is the makefiles that search for the .a to do dependency checking. By default the linker will against the shared library. Regards, Ryan Mansfield

Ryan Mansfield 2008-12-16T14:21:10Z Aaron Cripps http://community.qnx.com/sf/go/post18688 2008-12-16T13:59:11Z 2008-12-16T13:59:11Z

sorry, I may have missed one thing earlier ... if you have any extra options, you will want to add them as a flag under CCOPTS so,: LIBS += socket CCOPTS += -Bdynamic will use the -B flag on the command line, and include socket as a lib against which to link. -Aaron On Tuesday 16 December 2008 00:54:20 soujanya Yelchuri wrote: > Hi, > I want to use the io-pkt instead of io-net network stack for my project for > that purpose i am migrating my project to QNX 640. > > I am facing the isses in compilation on QNX 640 as it is using gcc 4.2.4. > I have not faced any problems when i migrated to QNX 632.(using gcc > 3.3.5).My project is developed on PPC target MPC 8270. > > I am able to compile the modules which does not have any dependencies and > which dependent on target libraries ex: socket But when i try to include > any library dynamically with an option > -Bdynamic to LIBS+= path in common.mk its displaying the flowing error > message. > > cpp > make[2]: *** No rule to make target `lib-Bdymanic.a', > > if i include statically with -Bstatic its displaying the flowing error > message. > > cpp > make[2]: *** No rule to make target `lib-static.a'; > > I did not face these kind of problems when i migrated to QNX 632. > I did not change my common.mk file.By defalut it should support the > previous verion of tool chain. > > On my PC only QNX 640 is running. > I am sure about the QNX_HOST and the target path. > > One more isse what iam facing on QNX 640 is, > If a libarary which is included without any option in LIBS+= path in > common.mk the compiler is searching for its .a. But by default it should > search for .so in the included paths. > I included the path of the libraries but its searching for its .a > Why is it so? > > Why same common.mk file which is supported in QNX 632 not getting supported > in QNX 640???? > > Regards, > Soujanya. > > > > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post18672

Aaron Cripps 2008-12-16T13:59:11Z Aaron Cripps http://community.qnx.com/sf/go/post18687 2008-12-16T13:56:24Z 2008-12-16T13:56:24Z

you want to do: LIBS += libname not LIBS += -Blibname Hope this helps, -Aaron On Tuesday 16 December 2008 00:54:20 soujanya Yelchuri wrote: > Hi, > I want to use the io-pkt instead of io-net network stack for my project for > that purpose i am migrating my project to QNX 640. > > I am facing the isses in compilation on QNX 640 as it is using gcc 4.2.4. > I have not faced any problems when i migrated to QNX 632.(using gcc > 3.3.5).My project is developed on PPC target MPC 8270. > > I am able to compile the modules which does not have any dependencies and > which dependent on target libraries ex: socket But when i try to include > any library dynamically with an option > -Bdynamic to LIBS+= path in common.mk its displaying the flowing error > message. > > cpp > make[2]: *** No rule to make target `lib-Bdymanic.a', > > if i include statically with -Bstatic its displaying the flowing error > message. > > cpp > make[2]: *** No rule to make target `lib-static.a'; > > I did not face these kind of problems when i migrated to QNX 632. > I did not change my common.mk file.By defalut it should support the > previous verion of tool chain. > > On my PC only QNX 640 is running. > I am sure about the QNX_HOST and the target path. > > One more isse what iam facing on QNX 640 is, > If a libarary which is included without any option in LIBS+= path in > common.mk the compiler is searching for its .a. But by default it should > search for .so in the included paths. > I included the path of the libraries but its searching for its .a > Why is it so? > > Why same common.mk file which is supported in QNX 632 not getting supported > in QNX 640???? > > Regards, > Soujanya. > > > > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post18672

Aaron Cripps 2008-12-16T13:56:24Z soujanya Yelchuri http://community.qnx.com/sf/go/post18672 2008-12-16T05:54:16Z 2008-12-16T05:54:16Z

Hi, I want to use the io-pkt instead of io-net network stack for my project for that purpose i am migrating my project to QNX 640. I am facing the isses in compilation on QNX 640 as it is using gcc 4.2.4. I have not faced any problems when i migrated to QNX 632.(using gcc 3.3.5).My project is developed on PPC target MPC 8270. I am able to compile the modules which does not have any dependencies and which dependent on target libraries ex: socket But when i try to include any library dynamically with an option -Bdynamic to LIBS+= path in common.mk its displaying the flowing error message. cpp make[2]: *** No rule to make target `lib-Bdymanic.a', if i include statically with -Bstatic its displaying the flowing error message. cpp make[2]: *** No rule to make target `lib-static.a'; I did not face these kind of problems when i migrated to QNX 632. I did not change my common.mk file.By defalut it should support the previous verion of tool chain. On my PC only QNX 640 is running. I am sure about the QNX_HOST and the target path. One more isse what iam facing on QNX 640 is, If a libarary which is included without any option in LIBS+= path in common.mk the compiler is searching for its .a. But by default it should search for .so in the included paths. I included the path of the libraries but its searching for its .a Why is it so? Why same common.mk file which is supported in QNX 632 not getting supported in QNX 640???? Regards, Soujanya.

soujanya Yelchuri 2008-12-16T05:54:16Z Warren Deitch http://community.qnx.com/sf/go/post18665 2008-12-15T23:00:24Z 2008-12-15T23:00:24Z

Yes, it is Dell's latest offering - which was why I included all the descriptions. Being a quad core with lots of ram was a powerfull inducement for me as a lot of my work is CPU bound on lesser machines. The peripherals chipset is ICH9 which already has some/all? support. The hard disk SATA ahci is a supported standard as well. Networking started but then stopped? I know this is a new flavour of OS (in this case also on new hardware) and I am willing to help resolve the 'potential features' (bugs). My previous laptop was a Dell Inspiron 8600 and after a while, with the assistance of the QNX community, the only driver I ended up was missing was video - vesa was an adequate workaround. sloginfo is showing almost continuous :- Dec 16 07:55:41 2 19 0 eide_reset: complete 302 ms Dec 16 07:55:41 2 19 0 eide_init_devices: ST9320421ASG path 2, tid 0, udma -1, mdma -1, sdma -1, pio 4, mblk 16 Dec 16 07:55:41 2 19 0 eide_reset: complete 301 ms Dec 16 07:55:41 2 19 0 eide_init_devices: ST9320421ASG path 2, tid 0, udma -1, mdma -1, sdma -1, pio 4, mblk 16 Dec 16 07:55:41 2 19 0 eide_reset: complete 301 ms Dec 16 07:55:41 2 19 0 eide_init_devices: ST9320421ASG path 2, tid 0, udma -1, mdma -1, sdma -1, pio 4, mblk 16 Warren

Warren Deitch 2008-12-15T23:00:24Z Mike Lemke http://community.qnx.com/sf/go/post18628 2008-12-15T18:05:19Z 2008-12-15T18:05:19Z

Hi Thomas - SDP issues such as missing links :( can go into the qnx_momentics_community_support forum. (We're not going to change the name of the momentics forum, what with links, aliases, etc. We're also not going to create a 'successor' forum, at least not at this time.) As per usual, any technology-specific issues (networking, kernel, tools, etc.) should be posted in the technology projects that already exist. Thanks! Mike

Mike Lemke 2008-12-15T18:05:19Z Thomas Fletcher http://community.qnx.com/sf/go/post18607 2008-12-15T15:14:49Z 2008-12-15T15:14:49Z

Just wondering where you would like to see posts about general 6.4.0 Post-Release issues that are being encountered. They aren't specific Tools/OS issues but rather issues with how SDP is set-up/configured (like missing links etc) Thanks, Thomas

Thomas Fletcher 2008-12-15T15:14:49Z Malte Mundt http://community.qnx.com/sf/go/post18582 2008-12-15T10:33:05Z 2008-12-15T10:33:05Z

Great report! Could it be that this laptop is brand new? I don't think it is expected to work. For hard disk operation to be fast, e.g. a corresponding driver is needed. For QNX to develop a driver, first the documentation is needed. And also a developer is needed who has time to do it. :-) Using QNX Neutrino natively on laptops isn't the primary purpose of this embedding OS. However, QNX does support a reference laptop which currently is the Dell Latitude D830. The next one will be the E5500 as far as I heard. Cheers - Malte

Malte Mundt 2008-12-15T10:33:05Z Warren Deitch http://community.qnx.com/sf/go/post18570 2008-12-13T12:24:15Z 2008-12-13T12:24:15Z

I am trying to install 6.4.0 (qnxsdp-6.4.0-200810211530-nto.iso) onto a Dell M6400 with 8GB ram and two hard disks (7200 rpm). The success rate has been very low. To keep it simple, I am only trying to install onto the second drive (320GB) without installing the boot loader onto the first drive (250GB). Since the Windoze OS is XP x64 - all my partitioning tools will not install so I am not yet able to reduce the partition size on the first hard disk (nor can I find an anti-virus package for x64 - but that is another story) So far : If the hard disks are configured as SATA ahci - they are seen by the loader but install does not see/use the second drive. With the BIOS set to ATA mode both drives are usable. Install with type 179 took over 14 hours (Started about 1200 and went to bed about 0200 next morning). The QNX6 file system preparation was still going at 0200. Next day after I got the prompts for the file system, the copies from /cd to /hardisk took over 5 hours. It reminded me of the very early QNX4 system which let the CD shutdown between reads. The HD activity led was flashing about once per second. Because of other bad results (see below), a couple of days later I did a re-install using type 79 (QNX4). The only improvement was that the file-system preparation type was insignificant - again - I went to bed before it finished and again - the copies took more than 5 hours. Trying a 6.3.2 install was stopped because /dev/pci was never created. I have included the pci -vv and usb -vvvvvvvvVt reports (gzipped) Summary of problems encountered : 2 hard disks in SATA ahci mode does not appear to work. The second drive is seen while scanning but cannot be accessed - see below for a dump. Startup - On reboot the system is reported as being up for 49709 days 20 hours (-10 hours years treated as unsigned could be 136 years??). Since I MUST use windoze, the RTC is +10 hours not UTC (AU timezone). I think I have already seen this reported. Only 4 Gb of ram (out of 8 Gb) is reported or available for use. Video - although vesabios is reported as being available the ONLY mode that works is svga 640 x 480. Attempting to change to anything else vesabios or 1024 x 768 etc results in the screen becoming blank and a beep after approx 30 seconds (max 3 times). Only option is the power switch. BTW - will a decent nvidia driver become available ? Hard disk activity seems to be VERY slow. Both QNX6 and QNX4 filesystems. When shutting down - the HD activity LED flashes 1/sec for about 30 secs. Copying large files is hopeless. About 5 Mb / minute hard disk to same hard disk and about 3 Mb / minute hard disk to a USB 2.0 drive. The system monitor does not even move off the stop. Networking uses the Broadcom BCM5761e Gigabit PCIe (vid=14e4,did=1680) I added it to /etc/system/enum/devices/net for the tigon3 driver. On startup I get assigned an IP address via dhcp so I know the hardware works but that is it. I cannot ping anybody nor does qnet work, nor can other machines ping this host. Minor problem - USB drives - plug in a single usb drive (7 different types tried). They each mount as /fs/usb0 but if the usb drive is present at bootup it is mounted as /fs/hd10-dos-1. They all work OK - it just the inconsistent mount point. As expected, when plugged in while system as on, /fs/usb0, /fs/usb1, etc. At startup /fs/hd10-dos-1 /fs/hd20-dos-1 etc. The order at startup is consistent for the same drives over multiple startups - thanks. Minor report - while doing a pidin mem - the io-display and io-graphics have non-ASCII chars in the report (line starts with PCI vid,did) - see pidin.mem.gz After installing onto the second drive while in ATA mode, I then switched the BIOS back to SATA ahci mode. It seems to load .boot than fail after that. Press F1-F4 to select drive or select partition 1? 1 Hit Esc for .altboot............................................................................... Press the space bar to input boot options or D to disable DMA... Trying SATA ahci. Scanning for devices... ...found 2 devices Trying EIDE. Scanning for devices... <- what about the CD which is EIDE as reported in ATA mode Trying USB. Scanning for devices... Trying SATA mvSata. Scanning for devices... /.diskroot file not found on any filesysem. Starting with safe mode Spawn of /bin/sh failed : No such file or directory Warren

Warren Deitch 2008-12-13T12:24:15Z Gary Faulkner http://community.qnx.com/sf/go/post18331 2008-12-10T16:48:12Z 2008-12-10T16:48:12Z

Thanks again Robert. This will get me going!

Gary Faulkner 2008-12-10T16:48:12Z Robert Craig http://community.qnx.com/sf/go/post18329 2008-12-10T16:38:52Z 2008-12-10T16:38:52Z

Absolutely. It's branches/BC640_2406_release. This is actually the maintenance branch for the release so you'll also find that fixes that we think seriously immpact functionality are also added in, but they in no way break compatibility. I can also let you know the tag that was used to create the binary in the release, but I'd recommend using the newest source in that branch. -----Original Message----- From: Gary Faulkner [mailto:community-noreply@qnx.com] Sent: Wednesday, December 10, 2008 11:35 AM To: sdp640prerelease-community Subject: Re: RE: taking existing io-pkt drivers and running them under the released 6.4 version of io-pkt Thanks Robert. Is there a specific io-pkt source svn tag that represents the 6.4 release? _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post18327

Robert Craig 2008-12-10T16:38:52Z Gary Faulkner http://community.qnx.com/sf/go/post18327 2008-12-10T16:34:55Z 2008-12-10T16:34:55Z

Thanks Robert. Is there a specific io-pkt source svn tag that represents the 6.4 release?

Gary Faulkner 2008-12-10T16:34:55Z Robert Craig http://community.qnx.com/sf/go/post18322 2008-12-10T16:05:23Z 2008-12-10T16:05:23Z

Hi Gary: There were some stack internals that changed towards the end of the release cycle that necessitated a re-build of all drivers. This is something that we will be doing our utmost to avoid in the future. I suspect that you'll find that your driver also has to match the correct 6.3.2 code base as well. There are also some 6.3.2 vs. 6.4 reasons (most notably libc.so.2 vs. libc.so.3) that can impact things (although I suspect that a 6.3.2 built driver built against the proper code base will also work on 6.4 as long as you include libc.so.2 in your image) With that being said, we have a specific requirement for the next release to allow a driver to be built without having to download the entire source base. The tricky thing is the cascading dependency of headers which results in some 60 headers from the io-pkt tree having to be pulled in. We're working on figuring out how to reduce that number and hopefully include them as standard OS source headers. This way, at worst case, you will have to build the driver on the version of the OS that you plan to run it on. Hope that helps, Robert. -----Original Message----- From: Gary Faulkner [mailto:community-noreply@qnx.com] Sent: Wednesday, December 10, 2008 10:36 AM To: sdp640prerelease-community Subject: taking existing io-pkt drivers and running them under the released 6.4 version of io-pkt Hello: I asked this question here a while ago (http://community.qnx.com/sf/go/projects.community/discussion.qnx_softwa re_development_platfor.topc4692). Upon further reflection, I don't think I asked the correct question. I have a network driver that I wrote/ported to io-pkt under 6.3.2. I want to run it under io-pkt on 6.4. I'd really prefer to use the released version of io-pkt that comes with 6.4; however, I can't seem to find a way to do that. It seems like I must build my driver with the version of io-pkt that I run. With that being said, is there a way, or going to be a way, to allow for io-pkt driver development to be based against the released io-pkt in 6.4? or will we always need to download, build, and use a source version of io-pkt from subversion if we have custom drivers? Thanks!! -garyf _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post18306

Robert Craig 2008-12-10T16:05:23Z Gary Faulkner http://community.qnx.com/sf/go/post18306 2008-12-10T15:35:50Z 2008-12-10T15:35:50Z

Hello: I asked this question here a while ago (http://community.qnx.com/sf/go/projects.community/discussion.qnx_software_development_platfor.topc4692). Upon further reflection, I don't think I asked the correct question. I have a network driver that I wrote/ported to io-pkt under 6.3.2. I want to run it under io-pkt on 6.4. I'd really prefer to use the released version of io-pkt that comes with 6.4; however, I can't seem to find a way to do that. It seems like I must build my driver with the version of io-pkt that I run. With that being said, is there a way, or going to be a way, to allow for io-pkt driver development to be based against the released io-pkt in 6.4? or will we always need to download, build, and use a source version of io-pkt from subversion if we have custom drivers? Thanks!! -garyf

Gary Faulkner 2008-12-10T15:35:50Z Derek Leach http://community.qnx.com/sf/go/post18123 2008-12-08T15:29:09Z 2008-12-08T15:29:09Z

Please post the sloginfo as Misha suggested. Some hint: 1) Your Windows environment must be in 32-bit mode. 2) Try editting the display.conf and set the resolution to 800x600. 3) See if there is a vmware.log file - not sure if this will help, but I seen reference to this in google. -Derek

Derek Leach 2008-12-08T15:29:09Z Misha Nefedov http://community.qnx.com/sf/go/post18102 2008-12-08T13:49:39Z 2008-12-08T13:49:39Z

Could you please post the output of these commands (or just "relavent" portions of it): sloginfo ls -l /dev/io-display pci -v #VGA adapter part...

Misha Nefedov 2008-12-08T13:49:39Z Bill Houglum http://community.qnx.com/sf/go/post18066 2008-12-06T00:54:28Z 2008-12-06T00:54:28Z

VMware Server, Version 2.0.0, Build 116503 The /etc/system/config/display.conf file appears correct compared to versions suggested in other posts on the forum.

Bill Houglum 2008-12-06T00:54:28Z Sean Boudreau http://community.qnx.com/sf/go/post17599 2008-11-30T14:40:43Z 2008-11-30T14:40:43Z

On Sun, Nov 30, 2008 at 09:08:43AM -0500, Armin Steinhoff wrote: > > On Sat, Nov 29, 2008 at 07:24:18AM -0500, Armin Steinhoff wrote: > [clip] > > > > This is PR 39687 and is fixed in the 6.4 select(). > > > > -seanb > > libc.so.2 is part of QNX 6.4 ... do we have a chance to get this patch also in libc.so.2 ?? In 6.4 libc.so.2 only contains some gcc wrapper funcs. libc.so.2 has a dependency on libc.so.3 so the majority of functionality comes from there. So technically, yes libc.so.2 has the fix in 6.4 (via libc.so.3). > > In general ... some projects compiled with gcc 3.3.5 / QNX6.3.2 can't be succesfully recompiled with gcc 4.2.4/ QNX6.4. > > For instance ... I have ported (or recompiled) freetype-2.3.2 with gcc 3.3.5 /QNX6.3.2. When I copy this project to my QNX6.4 machine and start the recompilation with 'make clean, make' the compilation fails! > > Is the compile environment of QNX6.3.2 and QNX 6.4 incompatible ?? Not that I know of but specific questions can be asked in the tools forum. Regards, -seanb

Sean Boudreau 2008-11-30T14:40:43Z Armin Steinhoff http://community.qnx.com/sf/go/post17598 2008-11-30T14:08:38Z 2008-11-30T14:08:38Z

> On Sat, Nov 29, 2008 at 07:24:18AM -0500, Armin Steinhoff wrote: [clip] > > This is PR 39687 and is fixed in the 6.4 select(). > > -seanb libc.so.2 is part of QNX 6.4 ... do we have a chance to get this patch also in libc.so.2 ?? In general ... some projects compiled with gcc 3.3.5 / QNX6.3.2 can't be succesfully recompiled with gcc 4.2.4/ QNX6.4. For instance ... I have ported (or recompiled) freetype-2.3.2 with gcc 3.3.5 /QNX6.3.2. When I copy this project to my QNX6.4 machine and start the recompilation with 'make clean, make' the compilation fails! Is the compile environment of QNX6.3.2 and QNX 6.4 incompatible ?? Regards --Armin

Armin Steinhoff 2008-11-30T14:08:38Z Sean Boudreau http://community.qnx.com/sf/go/post17592 2008-11-29T13:34:56Z 2008-11-29T13:34:56Z

On Sat, Nov 29, 2008 at 07:24:18AM -0500, Armin Steinhoff wrote: > > On Fri, Nov 28, 2008 at 03:16:37PM -0500, Armin Steinhoff wrote: > > > > Hi Armin: > > > > Is it possible that the application was still running and > > > > driving io-pkt or did it look like it was stuck in a loop? > > > > > > No, it waits with a select statement in a loop ... the timeout val is set to > > zero, that means all components of the struct timeval are set to zero. No > > error status from the select call. > > > > So a non blocking select() ? You're using the 6.4 libc? > > No, our target application is linked statically agains the 6.3.2 libc. > .... and that's the point! I will recompile the target and link it against the 6.4 libc. I will keep you informed about the test result .. > This is PR 39687 and is fixed in the 6.4 select(). -seanb

Sean Boudreau 2008-11-29T13:34:56Z Armin Steinhoff http://community.qnx.com/sf/go/post17591 2008-11-29T12:24:15Z 2008-11-29T12:24:15Z

> On Fri, Nov 28, 2008 at 03:16:37PM -0500, Armin Steinhoff wrote: > > > Hi Armin: > > > Is it possible that the application was still running and > > > driving io-pkt or did it look like it was stuck in a loop? > > > > No, it waits with a select statement in a loop ... the timeout val is set to > zero, that means all components of the struct timeval are set to zero. No > error status from the select call. > > So a non blocking select() ? You're using the 6.4 libc? No, our target application is linked statically agains the 6.3.2 libc. .... and that's the point! I will recompile the target and link it against the 6.4 libc. I will keep you informed about the test result .. Best Regards --Armin > > -seanb

Armin Steinhoff 2008-11-29T12:24:15Z Sean Boudreau http://community.qnx.com/sf/go/post17577 2008-11-28T20:24:29Z 2008-11-28T20:24:29Z

On Fri, Nov 28, 2008 at 03:16:37PM -0500, Armin Steinhoff wrote: > > Hi Armin: > > Is it possible that the application was still running and > > driving io-pkt or did it look like it was stuck in a loop? > > No, it waits with a select statement in a loop ... the timeout val is set to zero, that means all components of the struct timeval are set to zero. No error status from the select call. So a non blocking select() ? You're using the 6.4 libc? -seanb

Sean Boudreau 2008-11-28T20:24:29Z Armin Steinhoff http://community.qnx.com/sf/go/post17576 2008-11-28T20:16:35Z 2008-11-28T20:16:35Z

> Hi Armin: > Is it possible that the application was still running and > driving io-pkt or did it look like it was stuck in a loop? No, it waits with a select statement in a loop ... the timeout val is set to zero, that means all components of the struct timeval are set to zero. No error status from the select call. > It's also > possible that there was a driver bug that resulted in constant looping, > but I haven't found anything internally in our data base on the 8139. I'm was using the official driver for the 8139. Best Regards --Armin > > R. > > -----Original Message----- > From: Armin Steinhoff [mailto:community-noreply@qnx.com] > Sent: Friday, November 28, 2008 12:57 PM > To: sdp640prerelease-community > Subject: Re: RE: io_packet consumes 100% CPU time > > > Hello Robert, > > the application was an UDP communication between a MS-Windows XP machine > and a QNX6.4 SMP machine. For the communication has been used an > additional installed Realtek 8139 PCI board. An onboard Gigabit adapter > was inactive and was not connected to the network ... but it was > recognized at startup of the QNX machine (en1). No switch has been used > ... both system was linked by a crossover patch cable. > > Best Regards > > --Armin > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post17562

Armin Steinhoff 2008-11-28T20:16:35Z Sean Boudreau http://community.qnx.com/sf/go/post17574 2008-11-28T20:06:41Z 2008-11-28T20:06:41Z

On Fri, Nov 28, 2008 at 03:04:20PM -0500, Armin Steinhoff wrote: > > Hello Sean, > Hello Robert, > > io_packet is sharing the IRQ 0xb with one of the USB channels ... but USB devices has not been used on this system. The Realtek PCI board was the only additional installed hardware. Does it happen if you slay off io-usb? -seanb

Sean Boudreau 2008-11-28T20:06:41Z Armin Steinhoff http://community.qnx.com/sf/go/post17573 2008-11-28T20:04:17Z 2008-11-28T20:04:17Z

Hello Sean, Hello Robert, io_packet is sharing the IRQ 0xb with one of the USB channels ... but USB devices has not been used on this system. The Realtek PCI board was the only additional installed hardware. Best Regards --Armin

Armin Steinhoff 2008-11-28T20:04:17Z Robert Craig http://community.qnx.com/sf/go/post17564 2008-11-28T18:08:56Z 2008-11-28T18:08:56Z

Hi Armin: Is it possible that the application was still running and driving io-pkt or did it look like it was stuck in a loop? It's also possible that there was a driver bug that resulted in constant looping, but I haven't found anything internally in our data base on the 8139. R. -----Original Message----- From: Armin Steinhoff [mailto:community-noreply@qnx.com] Sent: Friday, November 28, 2008 12:57 PM To: sdp640prerelease-community Subject: Re: RE: io_packet consumes 100% CPU time Hello Robert, the application was an UDP communication between a MS-Windows XP machine and a QNX6.4 SMP machine. For the communication has been used an additional installed Realtek 8139 PCI board. An onboard Gigabit adapter was inactive and was not connected to the network ... but it was recognized at startup of the QNX machine (en1). No switch has been used ... both system was linked by a crossover patch cable. Best Regards --Armin _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post17562

Robert Craig 2008-11-28T18:08:56Z Sean Boudreau http://community.qnx.com/sf/go/post17563 2008-11-28T17:59:54Z 2008-11-28T17:59:54Z

On Fri, Nov 28, 2008 at 12:56:48PM -0500, Armin Steinhoff wrote: > > Hello Robert, > > the application was an UDP communication between a MS-Windows XP machine and a QNX6.4 SMP machine. For the communication has been used an additional installed Realtek 8139 PCI board. An onboard Gigabit adapter was inactive and was not connected to the network ... but it was recognized at startup of the QNX machine (en1). No switch has been used ... both system was linked by a crossover patch cable. > Is io-pkt sharing an irq with anything (pidin irq)? -seanb

Sean Boudreau 2008-11-28T17:59:54Z Armin Steinhoff http://community.qnx.com/sf/go/post17562 2008-11-28T17:56:43Z 2008-11-28T17:56:43Z

Hello Robert, the application was an UDP communication between a MS-Windows XP machine and a QNX6.4 SMP machine. For the communication has been used an additional installed Realtek 8139 PCI board. An onboard Gigabit adapter was inactive and was not connected to the network ... but it was recognized at startup of the QNX machine (en1). No switch has been used ... both system was linked by a crossover patch cable. Best Regards --Armin

Armin Steinhoff 2008-11-28T17:56:43Z Robert Craig http://community.qnx.com/sf/go/post17557 2008-11-28T16:38:50Z 2008-11-28T16:38:50Z

Hmmm.... Do you have any more information about how the network was set up, what type of drivers were being used etc.? Were any other applications using io-pkt at the time? This isn't something we've seen before. Robert. -----Original Message----- From: Armin Steinhoff [mailto:community-noreply@qnx.com] Sent: Friday, November 28, 2008 11:36 AM To: sdp640prerelease-community Subject: io_packet consumes 100% CPU time Hello, after several hours of a product demonstration at a fair, io_packet consumes 100% CPU time of one core of a dual core system. I did online debugging over an UDP link. io_packet continued consuming the CPU time after disconnecting the QNX system from the network. (installed has been the last version of QNX 6.4 with SMP support for a dual core x86 machine) Regards Armin Steinhoff _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post17554

Robert Craig 2008-11-28T16:38:50Z Armin Steinhoff http://community.qnx.com/sf/go/post17554 2008-11-28T16:36:19Z 2008-11-28T16:36:19Z

Hello, after several hours of a product demonstration at a fair, io_packet consumes 100% CPU time of one core of a dual core system. I did online debugging over an UDP link. io_packet continued consuming the CPU time after disconnecting the QNX system from the network. (installed has been the last version of QNX 6.4 with SMP support for a dual core x86 machine) Regards Armin Steinhoff

Armin Steinhoff 2008-11-28T16:36:19Z Derek Leach http://community.qnx.com/sf/go/post17174 2008-11-24T15:02:51Z 2008-11-24T15:02:51Z

> > Edit you display.conf to use the vesabios driver, and use argb8888. Also I > > > think you need 800x600. We have already fixed this on trunk. > > > The 200810211530 version of Neutrino 6.4.0 doesn't have this change. Should it > ? > > I had to do it to get photon to start under VMWare server 2.0.2. This thread is about VirtualPC. There were changes for vmware as well, but it is still using the devg-vmware.so . If forces it to 1024x768 with 32-bit resolution for vmware. These changes are on release and trunk.

Derek Leach 2008-11-24T15:02:51Z Tim Jenkins(deleted) http://community.qnx.com/sf/go/post17016 2008-11-20T22:10:48Z 2008-11-20T22:10:48Z

> Edit you display.conf to use the vesabios driver, and use argb8888. Also I > think you need 800x600. We have already fixed this on trunk. The 200810211530 version of Neutrino 6.4.0 doesn't have this change. Should it? I had to do it to get photon to start under VMWare server 2.0.2.

Tim Jenkins(deleted) 2008-11-20T22:10:48Z Melanie Blower http://community.qnx.com/sf/go/post16909 2008-11-19T17:41:59Z 2008-11-19T17:41:59Z

I had the same problem as in the original post. I am working within a secure network that didn't allow me access to SVN server. In my case, I was able to solve the problem by entering proxy information in this tab: New Repository Location>Proxy

Melanie Blower 2008-11-19T17:41:59Z Malte Mundt http://community.qnx.com/sf/go/post16790 2008-11-18T16:02:35Z 2008-11-18T16:02:35Z

I see the same error installing 6.4 GA on a Kontron nanoETXexpress. The only difference is that the number is not 444, but 443. It only comes up after pressing ENTER after the installation. - Malte

Malte Mundt 2008-11-18T16:02:35Z Maurizio Rossi http://community.qnx.com/sf/go/post16630 2008-11-17T07:53:40Z 2008-11-17T07:53:40Z

Thank you. Regards, Maurizio Rossi

Maurizio Rossi 2008-11-17T07:53:40Z Misha Nefedov http://community.qnx.com/sf/go/post16578 2008-11-14T13:47:09Z 2008-11-14T13:47:09Z

Thank you, The problem is that the Boot Time is after the current time! The phlogin2 never considered this possibility. I modified the code to display “Uptime: unable to determine” in this situation. Best regards, -Misha.

Misha Nefedov 2008-11-14T13:47:09Z Maurizio Rossi http://community.qnx.com/sf/go/post16559 2008-11-14T08:20:05Z 2008-11-14T08:20:05Z

Misha, attached to this message the requested informations. I think the problem is related to the timezone/daylight setting. Best regards, Maurizio Rossi

Maurizio Rossi 2008-11-14T08:20:05Z stefan thiel http://community.qnx.com/sf/go/post16507 2008-11-13T10:57:55Z 2008-11-13T10:57:55Z

Hi Keith, tried to install newest release on a real PC hardware. There the script error does not come up. bye, Stefan ------------------------------------------------------------------------------ Stefan Thiel Electronic Engineering Multitest elektronische Systeme GmbH Äussere Oberaustrasse 4 D-83026 Rosenheim s.thiel@multitest.com Tel.: 08031406-464 Web: www.multitest.com ------------------------------------------------------------------------------ Geschaeftsfuehrer: Dipl.Ing. MBA Karl H. Funke Registergericht Traunstein - HRB 12453 ------------------------------------------------------------------------------ Keith Russell <community-noreply@qnx.com> 12.11.2008 17:12 Bitte antworten an post16476@community.qnx.com An sdp640prerelease-community <post16476@community.qnx.com> Kopie Thema Re: RE: syntax error in cd installation script Hi, By any chance are you disconnecting the virtual drive before you shutdown? It looks like the ISO has become disconnected or similar. I assume you are using the ISO file directly. Thanks, Keith Russell _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post16476 _______________________________________________________________________ This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. This footnote also confirms that this email message has been swept by MessageLabs (http://www.messagelabs.com) for the presence of computer viruses. _______________________________________________________________________ _______________________________________________________________________ This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. This footnote also confirms that this email message has been swept by MessageLabs (http://www.messagelabs.com) for the presence of computer viruses. _______________________________________________________________________

stefan thiel 2008-11-13T10:57:55Z stefan thiel http://community.qnx.com/sf/go/post16506 2008-11-13T08:56:45Z 2008-11-13T08:56:45Z

Hi Keith, no, the iso is 100% still in the "drive". Over all I only wanted to give you a feedback, because I have seen this with the alpha release and with the current release. It causes no problem. Bye, Stefan ------------------------------------------------------------------------------ Stefan Thiel Electronic Engineering Multitest elektronische Systeme GmbH Äussere Oberaustrasse 4 D-83026 Rosenheim s.thiel@multitest.com Tel.: 08031406-464 Web: www.multitest.com ------------------------------------------------------------------------------ Geschaeftsfuehrer: Dipl.Ing. MBA Karl H. Funke Registergericht Traunstein - HRB 12453 ------------------------------------------------------------------------------ Keith Russell <community-noreply@qnx.com> 12.11.2008 17:12 Bitte antworten an post16476@community.qnx.com An sdp640prerelease-community <post16476@community.qnx.com> Kopie Thema Re: RE: syntax error in cd installation script Hi, By any chance are you disconnecting the virtual drive before you shutdown? It looks like the ISO has become disconnected or similar. I assume you are using the ISO file directly. Thanks, Keith Russell _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post16476 _______________________________________________________________________ This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. This footnote also confirms that this email message has been swept by MessageLabs (http://www.messagelabs.com) for the presence of computer viruses. _______________________________________________________________________ _______________________________________________________________________ This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. This footnote also confirms that this email message has been swept by MessageLabs (http://www.messagelabs.com) for the presence of computer viruses. _______________________________________________________________________

stefan thiel 2008-11-13T08:56:45Z Misha Nefedov http://community.qnx.com/sf/go/post16493 2008-11-12T20:43:29Z 2008-11-12T20:43:29Z

Incorrect up-time problem Maurizio, I am unable to reproduce this problem. My guess is that the system time was drastically changed after or during boot. If you are still able to reproduce this problem, could you please run these commands in a terminal: # pidin info # date I would like to see what they report. Best regards, -Misha.

Misha Nefedov 2008-11-12T20:43:29Z Keith Russell http://community.qnx.com/sf/go/post16476 2008-11-12T16:12:49Z 2008-11-12T16:12:49Z

Hi, By any chance are you disconnecting the virtual drive before you shutdown? It looks like the ISO has become disconnected or similar. I assume you are using the ISO file directly. Thanks, Keith Russell

Keith Russell 2008-11-12T16:12:49Z stefan thiel http://community.qnx.com/sf/go/post16447 2008-11-12T12:18:13Z 2008-11-12T12:18:13Z

today I installed the new 6.4.0 realease into my MS VirtualPC. qnxsdp-6.4.0-x86-200810211530-nto.iso please find attached screenshot with the error still existing

stefan thiel 2008-11-12T12:18:13Z Elena Laskavaia http://community.qnx.com/sf/go/post16433 2008-11-11T22:13:51Z 2008-11-11T22:13:51Z

No error in configuration is unrelated. It is because we don't include source plugins for CDT. Mario Charest wrote: > Checked for the fun of it and on windows it was disable for me. Also the Product Configuration shows and error for Eclipse C/C++ Developments Tools SDK 4.0.3.... Is that because I'm setup to use 6.3.2? > > -----Original Message----- > From: Dominic Rath [mailto:community-noreply@qnx.com] > Sent: November-11-08 3:50 PM > To: sdp640prerelease-community > Subject: Re: Helpviewer on Linux > > Got it working! > > The configuration (About QNX Momentics IDE -> Configuration Details) showed the following difference: > > Windows: > Configured sites: > file:/C:/QNX640/target/qnx6/usr/help/eclipse/ > platform:/base/ > > Linux: > Configured sites: > platform:/base/ > > Going to "Software Updates -> Manage Configuration" I found that /opt/qnx640/target/qnx6/usr/help/eclipse/ was disabled. > > Selecting "Enable" from the "Available Tasks" made the missing documentation appear in the helpviewer after a reboot. > > Thanks for your help! > > Best regards, > > Dominic > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post16429 > > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post16431 >

Elena Laskavaia 2008-11-11T22:13:51Z Mario Charest http://community.qnx.com/sf/go/post16431 2008-11-11T20:58:22Z 2008-11-11T20:58:22Z

Checked for the fun of it and on windows it was disable for me. Also the Product Configuration shows and error for Eclipse C/C++ Developments Tools SDK 4.0.3.... Is that because I'm setup to use 6.3.2? -----Original Message----- From: Dominic Rath [mailto:community-noreply@qnx.com] Sent: November-11-08 3:50 PM To: sdp640prerelease-community Subject: Re: Helpviewer on Linux Got it working! The configuration (About QNX Momentics IDE -> Configuration Details) showed the following difference: Windows: Configured sites: file:/C:/QNX640/target/qnx6/usr/help/eclipse/ platform:/base/ Linux: Configured sites: platform:/base/ Going to "Software Updates -> Manage Configuration" I found that /opt/qnx640/target/qnx6/usr/help/eclipse/ was disabled. Selecting "Enable" from the "Available Tasks" made the missing documentation appear in the helpviewer after a reboot. Thanks for your help! Best regards, Dominic _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post16429

Mario Charest 2008-11-11T20:58:22Z Dominic Rath http://community.qnx.com/sf/go/post16430 2008-11-11T20:51:00Z 2008-11-11T20:51:00Z

s/reboot/Momentics restart/

Dominic Rath 2008-11-11T20:51:00Z Dominic Rath http://community.qnx.com/sf/go/post16429 2008-11-11T20:50:12Z 2008-11-11T20:50:12Z

Got it working! The configuration (About QNX Momentics IDE -> Configuration Details) showed the following difference: Windows: Configured sites: file:/C:/QNX640/target/qnx6/usr/help/eclipse/ platform:/base/ Linux: Configured sites: platform:/base/ Going to "Software Updates -> Manage Configuration" I found that /opt/qnx640/target/qnx6/usr/help/eclipse/ was disabled. Selecting "Enable" from the "Available Tasks" made the missing documentation appear in the helpviewer after a reboot. Thanks for your help! Best regards, Dominic

Dominic Rath 2008-11-11T20:50:12Z Dominic Rath http://community.qnx.com/sf/go/post16428 2008-11-11T20:32:41Z 2008-11-11T20:32:41Z

Thanks for the tips, still no luck though. Setting the environment didn't help, it was like that already. The file you've mentioned looks good, too: vmaster@milliways:~$ cat /opt/qnx640/host/linux/x86/usr/qde/eclipse/links/com.qnx.tools.ide.doc.qnx.momentics.link path=/opt/qnx640/target/qnx6/usr/help The content of that directory is the same on Linux and on Windows.

Dominic Rath 2008-11-11T20:32:41Z Elena Laskavaia http://community.qnx.com/sf/go/post16427 2008-11-11T20:16:08Z 2008-11-11T20:16:08Z

location path for 6.4.0 doc files is here cat /opt/qnx640/host/linux/x86/usr/qde/eclipse/links/com.qnx.tools.ide.doc.qnx.momentics.link should be path=/opt/qnx640/target/qnx6/usr/help Dominic Rath wrote: > P.S.: I've just noticed that Momentics/Eclipse under Linux lists only three of the com.qnx.doc.* plugins while the Windows version has dozens of them. > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post16426 >

Elena Laskavaia 2008-11-11T20:16:08Z Dominic Rath http://community.qnx.com/sf/go/post16426 2008-11-11T20:14:04Z 2008-11-11T20:14:04Z

P.S.: I've just noticed that Momentics/Eclipse under Linux lists only three of the com.qnx.doc.* plugins while the Windows version has dozens of them.

Dominic Rath 2008-11-11T20:14:04Z Elena Laskavaia http://community.qnx.com/sf/go/post16425 2008-11-11T20:13:39Z 2008-11-11T20:13:39Z

When you run IDE 4.5 you have to switch your environment vars to point to QNX 6.4.0, otherwise it would probably use 6.3.2. i.e. MAKEFLAGS=-I/opt/qnx640/target/qnx6/usr/include QNX_CONFIGURATION=/etc/qnx QNX_HOST=/opt/qnx640/host/linux/x86 QNX_JAVAHOME=/opt/qnx640/_jvm QNX_TARGET=/opt/qnx640/target/qnx6 p1=`echo $PATH | sed -e "s?$QNX_HOST/usr/bin:??" -e "s?/etc/qnx/bin:??"` PATH=$QNX_HOST/usr/bin:$p1 #start ide now qde that should correctly target 6.4.0 and pick docs for 6.4.0 OS. Dominic Rath wrote: > Hi, > > I've installed QNX SDP 6.4.0 (final release) on a (K)Ubuntu 8.04 machine in parallel with QNX 6.3.2. Almost everything seems to be working fine, but the helpviewer is missing most of the documentation, see the attached screenshot. > > IDE 4.0.1 shows all the documentation just fine, even after installing QNX 6.4.0 in parallel, but it only includes the 6.3 docs. 6.4.0 on a windows host also has all the documentation. > > Is this a known issue? Any ideas how to fix this? > > Regards, > > Dominic > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post16423 > > > ------------------------------------------------------------------------ >

Elena Laskavaia 2008-11-11T20:13:39Z Dominic Rath http://community.qnx.com/sf/go/post16423 2008-11-11T20:05:05Z 2008-11-11T20:05:05Z

Hi, I've installed QNX SDP 6.4.0 (final release) on a (K)Ubuntu 8.04 machine in parallel with QNX 6.3.2. Almost everything seems to be working fine, but the helpviewer is missing most of the documentation, see the attached screenshot. IDE 4.0.1 shows all the documentation just fine, even after installing QNX 6.4.0 in parallel, but it only includes the 6.3 docs. 6.4.0 on a windows host also has all the documentation. Is this a known issue? Any ideas how to fix this? Regards, Dominic

Dominic Rath 2008-11-11T20:05:05Z Maurizio Rossi http://community.qnx.com/sf/go/post16287 2008-11-10T08:15:07Z 2008-11-10T08:15:07Z

Thank you. Regards, Maurizio Rossi

Maurizio Rossi 2008-11-10T08:15:07Z Peter Manson http://community.qnx.com/sf/go/post16223 2008-11-07T16:49:52Z 2008-11-07T16:49:52Z

Hi Maurizio, I logged these issues in our problem tracking system: PR 62789: Some boot images not visible in secondary boot loader on qnx6 fs PR 62790: Invalid uptime on login window PR 62791: devb-umass issues with USB floppy drives and we're investigating. Peter Manson Manager, Test Team

Peter Manson 2008-11-07T16:49:52Z Raja chandru http://community.qnx.com/sf/go/post16162 2008-11-07T11:52:17Z 2008-11-07T11:52:17Z

> Hello All, > Running regression in sqlite3.6.4 through tcl8.5.5 > > *) I have downloaded compiled and installed the tcl8.5.5 in qnx neutrino 6.4. > 0 > *) Done the same thing for sqlite8.5.5 in qnx neutrino > *) I am going to the path /home/sqlite-3.6.4/tests and running # tclsh all. > test > It’s giving the following error > Invalid command name "sqlite3_soft_heap_limit" > While executing > "sqlite3_soft_heap_limit $soft_limit" > (file "./tester.tcl" line 50) > invoked from within > "source $testdir/tester.tcl" > (file "delete.test" line 17) > Please help me to resolve it > > Thanks in advance > Rajachandru Hi , I have solved the above issue and now i am getting new problem while running the all.test i am getting the following error... Test file error: couldn't fork child process: not enough memory crash2.test Test file error: couldn't fork child process: not enough memory crash3.test Test file error: couldn't fork child process: not enough memory crash4.test Test file error: couldn't fork child process: not enough memory crash5.test Please help me to solve it Thank U Regards Chandru

Raja chandru 2008-11-07T11:52:17Z Raja chandru http://community.qnx.com/sf/go/post16101 2008-11-06T15:35:24Z 2008-11-06T15:35:24Z

Hello All, Running regression in sqlite3.6.4 through tcl8.5.5 *) I have downloaded compiled and installed the tcl8.5.5 in qnx neutrino 6.4.0 *) Done the same thing for sqlite8.5.5 in qnx neutrino *) I am going to the path /home/sqlite-3.6.4/tests and running # tclsh all.test It’s giving the following error Invalid command name "sqlite3_soft_heap_limit" While executing "sqlite3_soft_heap_limit $soft_limit" (file "./tester.tcl" line 50) invoked from within "source $testdir/tester.tcl" (file "delete.test" line 17) Please help me to resolve it Thanks in advance Rajachandru

Raja chandru 2008-11-06T15:35:24Z Maurizio Rossi http://community.qnx.com/sf/go/post16081 2008-11-06T14:24:11Z 2008-11-06T14:24:11Z

Hi, there are some glitches in Momentics 6.4.0 running natively on x86. Secondary boot loader on qnx6 fs. If the output of the secondary boot loader in on the bottom row of the screen only the first boot image can be seen. You can select any available boot image but only the first one is visible. Login window. After system boot the login window reports an invalid uptime (se the attached image in logwin.png). The reported value is correct after one hour of system uptime. devb-umass and USB floppy drives. Plugging in an USB floppy drive cause some strange things. devb-umass started automatically does not handle the device (see usbplug.txt). The utility usb -vv reports a valid USB floppy device connected to the USB interface. At this point the system runs slower and one thread in the "mcd" daemon stays in the "READY" state. Starting manually devb-umass cause the device to appear in "/dev" as "/dev/hd1". At this point "mount -t dos /dev/hd1 /myfd" reports success (exit code 0) but "ls -la /myfd" reports "No such device or address (/myfd)" (see usbfloppy.txt). Regards, Maurizio Rossi

Maurizio Rossi 2008-11-06T14:24:11Z Malte Mundt http://community.qnx.com/sf/go/post15902 2008-11-04T10:50:46Z 2008-11-04T10:50:46Z

Done - PR# 62665

Malte Mundt 2008-11-04T10:50:46Z Nicolas Gagnon http://community.qnx.com/sf/go/post15861 2008-11-03T16:34:05Z 2008-11-03T16:34:05Z

Malte, File a PR, since this is the best way to track the job to be completed. Nicolas -----Original Message----- From: Malte Mundt [mailto:community-noreply@qnx.com] Sent: Monday, November 03, 2008 11:23 AM To: sdp640prerelease-community Subject: Re: RE: devn-e1000.so Shall I file PR on this? Or can I assume an io-pkt version will be available with the next release or as a patch soon? _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post15860

Nicolas Gagnon 2008-11-03T16:34:05Z Malte Mundt http://community.qnx.com/sf/go/post15860 2008-11-03T16:23:16Z 2008-11-03T16:23:16Z

Shall I file PR on this? Or can I assume an io-pkt version will be available with the next release or as a patch soon?

Malte Mundt 2008-11-03T16:23:16Z Sean Boudreau http://community.qnx.com/sf/go/post15786 2008-10-31T13:00:40Z 2008-10-31T13:00:40Z

On Fri, Oct 31, 2008 at 08:30:53AM -0400, Matthias Klein wrote: > Hello Sean, > > On Fri, Oct 24, 2008 at 11:31:52AM -0400, Matthias Klein wrote: > > Hello, > > > > the NIC's inside the SIMATIC BOX PC 627B (6ES7647-6BA10-1AX0) is not working with the devp-i82544.so on QNX 6.4 M8. > > The old devn-i82544.so is working fine. > > > > > Should be fixed in the networking project repo. > > Today I tested with the release version of 6.4: It does NOT work. > > Should the fixes be inside the relase version, or do I have to complie the network repo. myself ? > The fix went in after 6.4 was set. It's fixed in the networking project repo. -seanb

Sean Boudreau 2008-10-31T13:00:40Z Matthias Klein http://community.qnx.com/sf/go/post15783 2008-10-31T12:30:51Z 2008-10-31T12:30:51Z

Hello Sean, On Fri, Oct 24, 2008 at 11:31:52AM -0400, Matthias Klein wrote: > Hello, > > the NIC's inside the SIMATIC BOX PC 627B (6ES7647-6BA10-1AX0) is not working with the devp-i82544.so on QNX 6.4 M8. > The old devn-i82544.so is working fine. > > Should be fixed in the networking project repo. Today I tested with the release version of 6.4: It does NOT work. Should the fixes be inside the relase version, or do I have to complie the network repo. myself ? We have the same problem also on the the SIMATIC Rack PC 847B (6ES7643-8AA13-0XX0). In the attachment is the output of pci -vvv on the Rack PC... Best regards Matthias

Matthias Klein 2008-10-31T12:30:51Z Malte Mundt http://community.qnx.com/sf/go/post15779 2008-10-31T10:08:05Z 2008-10-31T10:08:05Z

So I assume it will be part of the next release. Thanks! - Malte

Malte Mundt 2008-10-31T10:08:05Z Robert Craig http://community.qnx.com/sf/go/post15748 2008-10-30T18:44:06Z 2008-10-30T18:44:06Z

The e1000 device was a new io-net driver port of brand new Intel code that Hugh did and wasn't included in 6.4. Robert. -----Original Message----- From: Nicolas Gagnon [mailto:community-noreply@qnx.com] Sent: Thursday, October 30, 2008 1:16 PM To: sdp640prerelease-community Subject: RE: devn-e1000.so Hi Malte, I've looked at the shiplist for 6.4.0 and I couldn't find a trace of either devn-e1000.so or devnp-e1000.so. So I doubt they will be available on 6.4.0. You might want to talk with either Robert C. or Hugh about it. Nicolas -----Original Message----- From: Malte Mundt [mailto:community-noreply@qnx.com] Sent: Thursday, October 30, 2008 12:46 PM To: sdp640prerelease-community Subject: devn-e1000.so On the Kontron nanoETXexpress, I saw a driver called devn-e1000.so running. Will this driver - or an io-pkt equivalent - be part of the 6.4 release? Would make sense... or not? :) _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post15728 _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post15735

Robert Craig 2008-10-30T18:44:06Z Nicolas Gagnon http://community.qnx.com/sf/go/post15735 2008-10-30T17:16:08Z 2008-10-30T17:16:08Z

Hi Malte, I've looked at the shiplist for 6.4.0 and I couldn't find a trace of either devn-e1000.so or devnp-e1000.so. So I doubt they will be available on 6.4.0. You might want to talk with either Robert C. or Hugh about it. Nicolas -----Original Message----- From: Malte Mundt [mailto:community-noreply@qnx.com] Sent: Thursday, October 30, 2008 12:46 PM To: sdp640prerelease-community Subject: devn-e1000.so On the Kontron nanoETXexpress, I saw a driver called devn-e1000.so running. Will this driver - or an io-pkt equivalent - be part of the 6.4 release? Would make sense... or not? :) _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post15728

Nicolas Gagnon 2008-10-30T17:16:08Z Malte Mundt http://community.qnx.com/sf/go/post15728 2008-10-30T16:45:58Z 2008-10-30T16:45:58Z

On the Kontron nanoETXexpress, I saw a driver called devn-e1000.so running. Will this driver - or an io-pkt equivalent - be part of the 6.4 release? Would make sense... or not? :)

Malte Mundt 2008-10-30T16:45:58Z Aaron Cripps http://community.qnx.com/sf/go/post15658 2008-10-29T13:05:58Z 2008-10-29T13:05:58Z

I haven't tried that yet ... I'll do that tomorrow; I'm not in the office today, and I left the laptop there. When I know, you'll know :) -Aaron On October 28, 2008 21:07:59 Robert Craig wrote: > Hi Aaron: > > If you assign an IP address does it work? What does nicinfo > report for the interface? > > Robert. >

Aaron Cripps 2008-10-29T13:05:58Z Robert Craig http://community.qnx.com/sf/go/post15651 2008-10-29T01:07:56Z 2008-10-29T01:07:56Z

Hi Aaron: If you assign an IP address does it work? What does nicinfo report for the interface? Robert. -----Original Message----- From: Aaron Cripps [mailto:community-noreply@qnx.com] Sent: Friday, October 24, 2008 1:28 PM To: sdp640prerelease-community Subject: Re: devnp-i82544.so not working with device ID 0x109A Hey Robert, that fixes the problem I was seeing too (with devnp-speedo, and my laptop not booting up). unfortunately, now dhcp.client isn't supplying en0 with an IP. -Aaron On Friday 24 October 2008 11:36:01 Robert Craig wrote: > Hi Mathias: > > We'll have to take a look and see if we've got one of those > in-house. The devn drivers are indeed still included with the final > release for exactly this reason. The older driver has way more mileage > on it then the new ones so we wanted to make sure that there was a > backup shipped "just in case". All you need to do have the old driver > enumerate over the new is go into etc/system/enum/devices/net and change > "devnp-i82544" to "devn-i82544" (remove the p). > > Robert. > > -----Original Message----- > From: Matthias Klein [mailto:community-noreply@qnx.com] > Sent: Friday, October 24, 2008 11:32 AM > To: sdp640prerelease-community > Subject: devnp-i82544.so not working with device ID 0x109A > > Hello, > > the NIC's inside the SIMATIC BOX PC 627B (6ES7647-6BA10-1AX0) is not > working with the devp-i82544.so on QNX 6.4 M8. > The old devn-i82544.so is working fine. > > The new driver enmuerates the wrong media rate and does not receive any > packet: (See the log files in the attachment) > > old: > INTEL 82544 Gigabit (Copper) Ethernet Controller > > Physical Node ID ........................... 000E8C 9E6045 > Current Physical Node ID ................... 000E8C 9E6045 > Current Operation Rate ..................... 100.00 Mb/s full-duplex > Active Interface Type ...................... MII > Active PHY address ....................... 0 > Maximum Transmittable data Unit ............ 1514 > Maximum Receivable data Unit ............... 0 > Hardware Interrupt ......................... 0xb > Memory Aperture ............................ 0xfc900000 - 0xfc91ffff > Promiscuous Mode ........................... Off > Multicast Support .......................... Enabled > > > new: > INTEL 82544 Gigabit (Copper) Ethernet Controller > > Physical Node ID ........................... 000E8C 9E6045 > Current Physical Node ID ................... 000E8C 9E6045 > Current Operation Rate ..................... 1000.00 Mb/s full-duplex > Active Interface Type ...................... MII > Active PHY address ....................... 0 > Maximum Transmittable data Unit ............ 1500 > Maximum Receivable data Unit ............... 0 > Hardware Interrupt ......................... 0xb > Memory Aperture ............................ 0xfc900000 - 0xfc91ffff > Promiscuous Mode ........................... Off > Multicast Support .......................... Enabled > > > Will the old drivers be included in the final release of 6.4 ? > > > Best regards > > Matthias > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post15517 > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post15518 _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post15521

Robert Craig 2008-10-29T01:07:56Z Mike Lemke http://community.qnx.com/sf/go/post15608 2008-10-28T13:06:32Z 2008-10-28T13:06:32Z

Thank you for your participation in the SDP 6.4.0 pre-release forum - we'll be releasing the final version of SDP 6.4.0 very shortly. Work has begun towards our next release - availability of a pre-release on Foundry27 likely won't occur for a few months. Thanks Mike Lemke

Mike Lemke 2008-10-28T13:06:32Z Mike Lemke http://community.qnx.com/sf/go/post15607 2008-10-28T12:59:54Z 2008-10-28T12:59:54Z

Hello Matthias, Sorry for the delay - I've just reviewed the final shiplist and yes, the library is included in the final release for x86 targets. Thanks Mike -----Original Message----- From: Matthias Klein [mailto:community-noreply@qnx.com] Sent: Monday, October 27, 2008 12:01 PM To: sdp640prerelease-community Subject: Re: DDD using XPhoton * push * >The proper /lib/libstdc++.so.2.2.10.0 will be included in the next posted build. The library is not included in QNX6.4 M8. Will the library be in the final release ? _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post15567

Mike Lemke 2008-10-28T12:59:54Z Sean Boudreau http://community.qnx.com/sf/go/post15568 2008-10-27T16:04:12Z 2008-10-27T16:04:12Z

On Fri, Oct 24, 2008 at 11:31:52AM -0400, Matthias Klein wrote: > Hello, > > the NIC's inside the SIMATIC BOX PC 627B (6ES7647-6BA10-1AX0) is not working with the devp-i82544.so on QNX 6.4 M8. > The old devn-i82544.so is working fine. > Should be fixed in the networking project repo. -seanb

Sean Boudreau 2008-10-27T16:04:12Z Matthias Klein http://community.qnx.com/sf/go/post15567 2008-10-27T16:00:46Z 2008-10-27T16:00:46Z

* push * >The proper /lib/libstdc++.so.2.2.10.0 will be included in the next posted build. The library is not included in QNX6.4 M8. Will the library be in the final release ?

Matthias Klein 2008-10-27T16:00:46Z Sean Boudreau http://community.qnx.com/sf/go/post15551 2008-10-27T13:30:48Z 2008-10-27T13:30:48Z

On Sun, Oct 26, 2008 at 07:20:38PM -0400, Gary Faulkner wrote: > Thanks Sean! exactly what i was looking for. > > I do have a follow-on question though. When 6.4 releases, what is the procedure for incorporating customer-specific networking hardware drivers? still building against the io-pkt source tree? or something else? > I guess the ideal would be to bring it over to the io-pkt framework. Maybe set up a local sys/dev_<foo> in your networking tree. There's no hard and fast documented procedure. Regards, -seanb

Sean Boudreau 2008-10-27T13:30:48Z Gary Faulkner http://community.qnx.com/sf/go/post15543 2008-10-26T23:20:35Z 2008-10-26T23:20:35Z

Thanks Sean! exactly what i was looking for. I do have a follow-on question though. When 6.4 releases, what is the procedure for incorporating customer-specific networking hardware drivers? still building against the io-pkt source tree? or something else? Thanks!! -garyf

Gary Faulkner 2008-10-26T23:20:35Z Sean Boudreau http://community.qnx.com/sf/go/post15541 2008-10-26T16:20:07Z 2008-10-26T16:20:07Z

On Sun, Oct 26, 2008 at 12:12:44PM -0400, Gary Faulkner wrote: > Hello: > > I'm wondering what the correct approach for new network drivers in 6.4 is thought to be? > > So here's where I am; i have a driver that works under io-pkt under 6.3.2. To get here, i took the io-pkt source, added in my driver, and built it under the 6.3.2 toolchain. > > However, the io-pkt source (at least the last one I updated... probably about 2 months ago), wants to build armbe binaries when I do a "make CPULIST=arm", and the 6.4 toochain no longer has armbe support in it (not a huge deal for me, because we're running the imx31/armle). > > So... my questions are... first, should I be doing something different under 6.4, since io-pkt is integrated? I did a very cursory check online to try to find a pointer on this, and didn't find anything. > > Or, is there an update for the io-pkt source to remove the armbe building rules? > > Thanks!! For now, to build all supported variants you can do (for example): # make EXCLUDE_CPULIST="arm mips" install Then after that if you want armle, mipsbe: # make VARIANTLIST="le" install # make VARIANTLIST="be" install Or if you only want armle from the outset # make CPULIST=arm VARIANTLIST=le install Hope that helps. -seanb

Sean Boudreau 2008-10-26T16:20:07Z Gary Faulkner http://community.qnx.com/sf/go/post15540 2008-10-26T16:12:40Z 2008-10-26T16:12:40Z

Hello: I'm wondering what the correct approach for new network drivers in 6.4 is thought to be? So here's where I am; i have a driver that works under io-pkt under 6.3.2. To get here, i took the io-pkt source, added in my driver, and built it under the 6.3.2 toolchain. However, the io-pkt source (at least the last one I updated... probably about 2 months ago), wants to build armbe binaries when I do a "make CPULIST=arm", and the 6.4 toochain no longer has armbe support in it (not a huge deal for me, because we're running the imx31/armle). So... my questions are... first, should I be doing something different under 6.4, since io-pkt is integrated? I did a very cursory check online to try to find a pointer on this, and didn't find anything. Or, is there an update for the io-pkt source to remove the armbe building rules? Thanks!! -garyf

Gary Faulkner 2008-10-26T16:12:40Z Aaron Cripps http://community.qnx.com/sf/go/post15521 2008-10-24T17:28:00Z 2008-10-24T17:28:00Z

Hey Robert, that fixes the problem I was seeing too (with devnp-speedo, and my laptop not booting up). unfortunately, now dhcp.client isn't supplying en0 with an IP. -Aaron On Friday 24 October 2008 11:36:01 Robert Craig wrote: > Hi Mathias: > > We'll have to take a look and see if we've got one of those > in-house. The devn drivers are indeed still included with the final > release for exactly this reason. The older driver has way more mileage > on it then the new ones so we wanted to make sure that there was a > backup shipped "just in case". All you need to do have the old driver > enumerate over the new is go into etc/system/enum/devices/net and change > "devnp-i82544" to "devn-i82544" (remove the p). > > Robert. > > -----Original Message----- > From: Matthias Klein [mailto:community-noreply@qnx.com] > Sent: Friday, October 24, 2008 11:32 AM > To: sdp640prerelease-community > Subject: devnp-i82544.so not working with device ID 0x109A > > Hello, > > the NIC's inside the SIMATIC BOX PC 627B (6ES7647-6BA10-1AX0) is not > working with the devp-i82544.so on QNX 6.4 M8. > The old devn-i82544.so is working fine. > > The new driver enmuerates the wrong media rate and does not receive any > packet: (See the log files in the attachment) > > old: > INTEL 82544 Gigabit (Copper) Ethernet Controller > > Physical Node ID ........................... 000E8C 9E6045 > Current Physical Node ID ................... 000E8C 9E6045 > Current Operation Rate ..................... 100.00 Mb/s full-duplex > Active Interface Type ...................... MII > Active PHY address ....................... 0 > Maximum Transmittable data Unit ............ 1514 > Maximum Receivable data Unit ............... 0 > Hardware Interrupt ......................... 0xb > Memory Aperture ............................ 0xfc900000 - 0xfc91ffff > Promiscuous Mode ........................... Off > Multicast Support .......................... Enabled > > > new: > INTEL 82544 Gigabit (Copper) Ethernet Controller > > Physical Node ID ........................... 000E8C 9E6045 > Current Physical Node ID ................... 000E8C 9E6045 > Current Operation Rate ..................... 1000.00 Mb/s full-duplex > Active Interface Type ...................... MII > Active PHY address ....................... 0 > Maximum Transmittable data Unit ............ 1500 > Maximum Receivable data Unit ............... 0 > Hardware Interrupt ......................... 0xb > Memory Aperture ............................ 0xfc900000 - 0xfc91ffff > Promiscuous Mode ........................... Off > Multicast Support .......................... Enabled > > > Will the old drivers be included in the final release of 6.4 ? > > > Best regards > > Matthias > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post15517 > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post15518

Aaron Cripps 2008-10-24T17:28:00Z Robert Craig http://community.qnx.com/sf/go/post15518 2008-10-24T15:35:58Z 2008-10-24T15:35:58Z

Hi Mathias: We'll have to take a look and see if we've got one of those in-house. The devn drivers are indeed still included with the final release for exactly this reason. The older driver has way more mileage on it then the new ones so we wanted to make sure that there was a backup shipped "just in case". All you need to do have the old driver enumerate over the new is go into etc/system/enum/devices/net and change "devnp-i82544" to "devn-i82544" (remove the p). Robert. -----Original Message----- From: Matthias Klein [mailto:community-noreply@qnx.com] Sent: Friday, October 24, 2008 11:32 AM To: sdp640prerelease-community Subject: devnp-i82544.so not working with device ID 0x109A Hello, the NIC's inside the SIMATIC BOX PC 627B (6ES7647-6BA10-1AX0) is not working with the devp-i82544.so on QNX 6.4 M8. The old devn-i82544.so is working fine. The new driver enmuerates the wrong media rate and does not receive any packet: (See the log files in the attachment) old: INTEL 82544 Gigabit (Copper) Ethernet Controller Physical Node ID ........................... 000E8C 9E6045 Current Physical Node ID ................... 000E8C 9E6045 Current Operation Rate ..................... 100.00 Mb/s full-duplex Active Interface Type ...................... MII Active PHY address ....................... 0 Maximum Transmittable data Unit ............ 1514 Maximum Receivable data Unit ............... 0 Hardware Interrupt ......................... 0xb Memory Aperture ............................ 0xfc900000 - 0xfc91ffff Promiscuous Mode ........................... Off Multicast Support .......................... Enabled new: INTEL 82544 Gigabit (Copper) Ethernet Controller Physical Node ID ........................... 000E8C 9E6045 Current Physical Node ID ................... 000E8C 9E6045 Current Operation Rate ..................... 1000.00 Mb/s full-duplex Active Interface Type ...................... MII Active PHY address ....................... 0 Maximum Transmittable data Unit ............ 1500 Maximum Receivable data Unit ............... 0 Hardware Interrupt ......................... 0xb Memory Aperture ............................ 0xfc900000 - 0xfc91ffff Promiscuous Mode ........................... Off Multicast Support .......................... Enabled Will the old drivers be included in the final release of 6.4 ? Best regards Matthias _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post15517

Robert Craig 2008-10-24T15:35:58Z Matthias Klein http://community.qnx.com/sf/go/post15517 2008-10-24T15:31:45Z 2008-10-24T15:31:45Z

Hello, the NIC's inside the SIMATIC BOX PC 627B (6ES7647-6BA10-1AX0) is not working with the devp-i82544.so on QNX 6.4 M8. The old devn-i82544.so is working fine. The new driver enmuerates the wrong media rate and does not receive any packet: (See the log files in the attachment) old: INTEL 82544 Gigabit (Copper) Ethernet Controller Physical Node ID ........................... 000E8C 9E6045 Current Physical Node ID ................... 000E8C 9E6045 Current Operation Rate ..................... 100.00 Mb/s full-duplex Active Interface Type ...................... MII Active PHY address ....................... 0 Maximum Transmittable data Unit ............ 1514 Maximum Receivable data Unit ............... 0 Hardware Interrupt ......................... 0xb Memory Aperture ............................ 0xfc900000 - 0xfc91ffff Promiscuous Mode ........................... Off Multicast Support .......................... Enabled new: INTEL 82544 Gigabit (Copper) Ethernet Controller Physical Node ID ........................... 000E8C 9E6045 Current Physical Node ID ................... 000E8C 9E6045 Current Operation Rate ..................... 1000.00 Mb/s full-duplex Active Interface Type ...................... MII Active PHY address ....................... 0 Maximum Transmittable data Unit ............ 1500 Maximum Receivable data Unit ............... 0 Hardware Interrupt ......................... 0xb Memory Aperture ............................ 0xfc900000 - 0xfc91ffff Promiscuous Mode ........................... Off Multicast Support .......................... Enabled Will the old drivers be included in the final release of 6.4 ? Best regards Matthias

Matthias Klein 2008-10-24T15:31:45Z Neil Schellenberger http://community.qnx.com/sf/go/post15485 2008-10-23T21:39:45Z 2008-10-23T21:39:45Z

Oops. I just realized I conflated two different scenarios there. (Faulty wetware apparently...) Running the livecd directly (no VMware) on my laptop with an external montor causes the black screen and hang grief. I've never managed to get anything other than a photon error when running under VMware server. One hint I was given is that it may have something to do with the desktop colour depth, so maybe try fiddling with that? Regards, Neil

Neil Schellenberger 2008-10-23T21:39:45Z Neil Schellenberger http://community.qnx.com/sf/go/post15484 2008-10-23T21:33:58Z 2008-10-23T21:33:58Z

FWIW on my Dell Latitude D830 laptop running Fedora 9 and VMware Server 1.0.7 build-108231 I can only get Photon running on the laptop screen. Using an external monitor is tricky, at best. It usually starts up OK at 640x480 but any attempt to change causes a black screen and an apparent hang. On Thu, 2008-10-23 at 17:30 -0400, Matthias Klein wrote: > Hello Gary, > > I use QNX 6.4 M7 and QNX 6.4 M8 inside VMware Workstation 6.0.4 build-93057. > > I didn't have any problems with photon. Even 16:9 resolutions (e.g. 1680x1050) are working fine. > > > Best regards > > Matthias > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post15483 >

Neil Schellenberger 2008-10-23T21:33:58Z Matthias Klein http://community.qnx.com/sf/go/post15483 2008-10-23T21:30:28Z 2008-10-23T21:30:28Z

Hello Gary, I use QNX 6.4 M7 and QNX 6.4 M8 inside VMware Workstation 6.0.4 build-93057. I didn't have any problems with photon. Even 16:9 resolutions (e.g. 1680x1050) are working fine. Best regards Matthias

Matthias Klein 2008-10-23T21:30:28Z Dell Drummond(deleted) http://community.qnx.com/sf/go/post15479 2008-10-23T20:53:43Z 2008-10-23T20:53:43Z

I tried the live cd using vmplayer 2.0.4 build-93057 on my Linux box. Everything looks to be working so far, including USB - which is surprising since it is flaky with an xp guest. The obvious questions are: - do other guest OSes work correctly? - are the vm modules upto date with respect to the kernel?

Dell Drummond(deleted) 2008-10-23T20:53:43Z Gary Faulkner http://community.qnx.com/sf/go/post15462 2008-10-23T18:17:14Z 2008-10-23T18:17:14Z

I'm having a problem installing native 6.4 install into vmware. I've tried vmware server; install was successful, but photon could not be started vmware workstation 6.5; install was successful, but when photon started, i just got a blank screen vmware workstation 6.0.5; install was successful, but when photo started i just got a blank screen My question is: what version of vmware should i be using for this? If it matters, I'm running vmware on linux instead of windows. Thanks! -garyf

Gary Faulkner 2008-10-23T18:17:14Z Misha Nefedov http://community.qnx.com/sf/go/post15404 2008-10-22T14:07:19Z 2008-10-22T14:07:19Z

FYI: This fix is in the final release.

Misha Nefedov 2008-10-22T14:07:19Z Misha Nefedov http://community.qnx.com/sf/go/post15402 2008-10-22T14:01:12Z 2008-10-22T14:01:12Z

This was a bug in _cmdname(). It has been fixed. PR#61394. I believe you should be able to run your application as long as you use its absolute path when starting it.

Misha Nefedov 2008-10-22T14:01:12Z Matthias Klein http://community.qnx.com/sf/go/post15387 2008-10-22T09:05:29Z 2008-10-22T09:05:29Z

Hello, we have the same problem with the absolute path and the "Ap: Unable to open resource file" message. Is there a solution available ? Best regards Matthias

Matthias Klein 2008-10-22T09:05:29Z Martin Gagnon http://community.qnx.com/sf/go/post15383 2008-10-22T02:55:06Z 2008-10-22T02:55:06Z

> Here is the attachment. > > Regards, > > Ryan Mansfield It's work perfectly, thanks.. I've notice something else with application that is build in photon, the application must be started using an absolute path. starting it using a relative path (e.g: ./appname) doesn't work once the Phab project is compiled on QNX 6.40. I got the following error: Ap: Unable to open resource file (/net/qnx640_dev3.ndt.ca/home/NDT/phab_proj/x86/./phab_proj) You can try yourself with the small project I post on my previous post. -- Martin

Martin Gagnon 2008-10-22T02:55:06Z Ryan Mansfield http://community.qnx.com/sf/go/post15370 2008-10-21T19:12:48Z 2008-10-21T19:12:48Z

Here is the attachment. Regards, Ryan Mansfield

Ryan Mansfield 2008-10-21T19:12:48Z Ryan Mansfield http://community.qnx.com/sf/go/post15369 2008-10-21T19:12:06Z 2008-10-21T19:12:06Z

Please try the attached qphabtarg.mk. Install it into $QNX_TARGET/usr/include/mk/ Please let me know if you have any problems. Regards, Ryan Mansfield

Ryan Mansfield 2008-10-21T19:12:06Z Ryan Mansfield http://community.qnx.com/sf/go/post15356 2008-10-21T17:33:59Z 2008-10-21T17:33:59Z

Martin Gagnon wrote: > It does it with a brand new Phab project created in 6.4 with only the generated code from phab.. > > I really think there's something really broken here.. but I guess it's another bug... > > Here the Plain Phab project.. tell me if you have this "libang-c++.a" error like me... Yes, I can reproduce the problem. Regards, Ryan Mansfield

Ryan Mansfield 2008-10-21T17:33:59Z Martin Gagnon http://community.qnx.com/sf/go/post15353 2008-10-21T17:22:49Z 2008-10-21T17:22:49Z

It does it with a brand new Phab project created in 6.4 with only the generated code from phab.. I really think there's something really broken here.. but I guess it's another bug... Here the Plain Phab project.. tell me if you have this "libang-c++.a" error like me... Thanks -- Martin

Martin Gagnon 2008-10-21T17:22:49Z William Smith http://community.qnx.com/sf/go/post15309 2008-10-21T12:11:24Z 2008-10-21T12:11:24Z

I played with this script and incorporated parts into our scripts and it seems like a nice way of doing it.

William Smith 2008-10-21T12:11:24Z Martin Gagnon http://community.qnx.com/sf/go/post15308 2008-10-21T11:58:58Z 2008-10-21T11:58:58Z

I've compiled OpenSSH 5.1p1 from the pkgsrc tree and it work no problem.. Just follow the instruction here: http://community.qnx.com/sf/wiki/do/viewPage/projects.pkgsrc/wiki/Bootstrap But you have to skip a few steps on QNX 6.4: - ignore about the "Set up your kit" and "Optionally update your procnto" section (those apply to QNX 6.3.2 only) - Inside the "Set up your environment" section, ignore the change of the default compiler, only one is available on QNX 6.4, and it's more recent. - For the svn checkout, you just need to checkout HEAD_640... # svn checkout --username <username> http://community.qnx.com/svn/repos/pkgsrc/HEAD_640 You can test if it's work with the figlet program, as they show on the wiki. Then to compile openssh, just cd to security/openssh (from HEAD_640) and make it: # cd security/openssh # bmake install and that's it.. It may ask you to create a user and a group before it install it.. you do it so and re-type the "bmake install" command and it will complete the installation. -- Martin

Martin Gagnon 2008-10-21T11:58:58Z Martin Gagnon http://community.qnx.com/sf/go/post15244 2008-10-20T15:39:51Z 2008-10-20T15:39:51Z

I use the stock common.mk from the QNX 6.3 SP3 Phab project. -- Martin

Martin Gagnon 2008-10-20T15:39:51Z Ryan Mansfield http://community.qnx.com/sf/go/post15242 2008-10-20T15:19:09Z 2008-10-20T15:19:09Z

Colin Burgess wrote: > I suspect that the -lang-c++ is somehow getting into the LIBS variable? Definitely. But I can't see how LDPREF could be added to LIBS by the recursive makefiles. Probably being added in the project's common.mk. Regards, Ryan Mansfield

Ryan Mansfield 2008-10-20T15:19:09Z Colin Burgess http://community.qnx.com/sf/go/post15240 2008-10-20T15:03:58Z 2008-10-20T15:03:58Z

I suspect that the -lang-c++ is somehow getting into the LIBS variable? Ryan Mansfield wrote: > Martin Gagnon wrote: >> May it can help you but I have a similar problem with my projects but they are only simple >> Photon builder project (no eclipse). It's happens on C++ project when some file in the project are ".c". On those ".c" file we use: extern "C" { statements and it was compiling fine under QNX 6.3. >> >> Now in QNX 6.4, I get the following error: >> make[1]: *** No rule to make target `libang-c++.a', needed by `.....` >> >> It's seems that the option "-lang-c++" that QCC receive is passed to "gcc" during the linking and it make gcc to look for a "libang-c++.a" file. Pretty similar than what you get in your case with libBstatic' . > > The error you're getting is a make error. It is definitely a makefile > issue and not an issue with qcc/ld. > > Regards, > > Ryan Mansfield > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post15239 > -- cburgess@qnx.com

Colin Burgess 2008-10-20T15:03:58Z Ryan Mansfield http://community.qnx.com/sf/go/post15239 2008-10-20T15:02:21Z 2008-10-20T15:02:21Z

Martin Gagnon wrote: > May it can help you but I have a similar problem with my projects but they are only simple > Photon builder project (no eclipse). It's happens on C++ project when some file in the project are ".c". On those ".c" file we use: extern "C" { statements and it was compiling fine under QNX 6.3. > > Now in QNX 6.4, I get the following error: > make[1]: *** No rule to make target `libang-c++.a', needed by `.....` > > It's seems that the option "-lang-c++" that QCC receive is passed to "gcc" during the linking and it make gcc to look for a "libang-c++.a" file. Pretty similar than what you get in your case with libBstatic' . The error you're getting is a make error. It is definitely a makefile issue and not an issue with qcc/ld. Regards, Ryan Mansfield

Ryan Mansfield 2008-10-20T15:02:21Z Martin Gagnon http://community.qnx.com/sf/go/post15237 2008-10-20T14:52:35Z 2008-10-20T14:52:35Z

May it can help you but I have a similar problem with my projects but they are only simple Photon builder project (no eclipse). It's happens on C++ project when some file in the project are ".c". On those ".c" file we use: extern "C" { statements and it was compiling fine under QNX 6.3. Now in QNX 6.4, I get the following error: make[1]: *** No rule to make target `libang-c++.a', needed by `.....` It's seems that the option "-lang-c++" that QCC receive is passed to "gcc" during the linking and it make gcc to look for a "libang-c++.a" file. Pretty similar than what you get in your case with libBstatic' . I've try to make a new Eclipse project of type "Photon Appbuilder project" with my original Phab project file. So it used the Eclipse common.mk file, and I got exactly the same error. For my understanding, the problem is not in QDE (Eclipse) but more generally somewhere inside "/usr/qnx640/", may be related with QCC, or the make file tree... Thanks -- Martin

Martin Gagnon 2008-10-20T14:52:35Z Natalia Miyusova http://community.qnx.com/sf/go/post15149 2008-10-17T13:26:21Z 2008-10-17T13:26:21Z

Hi Radek, You mentioned that the QNX entry was removed from Add/Remove programs. How about uninstaller.exe, was it also removed from C:\QNX640\_uninstall\qnxsdp\6.4.0 folder? If you could attach uninstallation log file, it would help to investigate the problem. Log files (uninstall-opt-Log.txt or uninstal.txt) should be the %TEMP% folder. Natalia

Natalia Miyusova 2008-10-17T13:26:21Z Radek Pesina http://community.qnx.com/sf/go/post15122 2008-10-17T00:00:38Z 2008-10-17T00:00:38Z

Just letting you know that I tried to upgrade from QNX 6.4.0 Alpha M7 to Alpha M8 and upon using uninstall from add/remove programs for release M7, QNX disappeared from the list however was not uninstalled. Luckily this thread existed so I could get it going again (I too tried cleaning up the registry with no success). It appears there's some issues with uninstall for Alpha M7.

Radek Pesina 2008-10-17T00:00:38Z Derek Leach http://community.qnx.com/sf/go/post15116 2008-10-16T20:10:49Z 2008-10-16T20:10:49Z

we ran into this recently, modify the enmerator file to force the poulsbo driver to start in 800x600 ... notice -g option device(pci, ven=8086, dev=8108) # SGX/Poulsbo start/wait( dispconf -i "0x$(ven),0x$(dev),$(index)" -d poulsbo -g "xres=800,yres=600" ) requires/wait( $(IODISPLAY_CMD), ) mount(-Tio-display "vid=0x$(ven),did=0x$(dev),deviceindex=$(index)" ) the driver is reporting it supports all generic modes, and maybe it should, but currently all modes do not work, only 800x600, 1024x768, and 1280x1024

Derek Leach 2008-10-16T20:10:49Z Steve Reid http://community.qnx.com/sf/go/post15036 2008-10-15T16:42:34Z 2008-10-15T16:42:34Z

I'll make sure that the installation notes mention this. Steve Reid (stever@qnx.com) Technical Editor QNX Software Systems > -----Original Message----- > From: Keith Russell [mailto:community-noreply@qnx.com] > Sent: Wednesday, October 15, 2008 8:48 AM > To: sdp640prerelease-community > Subject: RE: No Docs / No Help > > Yes, this is by design. The documentation would take up too much room > uncompressed on a "live CD". > > The run from CD option is really to support the installation. It also > allows some level of non-destructive testing to ensure QNX will run on > your hardware. > > Cheers, > Keith Russell

Steve Reid 2008-10-15T16:42:34Z Martin Gagnon http://community.qnx.com/sf/go/post15016 2008-10-15T14:23:26Z 2008-10-15T14:23:26Z

Thank you very much, That's help a lot, it does exactly what we need. -- Martin

Martin Gagnon 2008-10-15T14:23:26Z Maurizio Rossi http://community.qnx.com/sf/go/post15008 2008-10-15T13:01:41Z 2008-10-15T13:01:41Z

Attached to this message some more informations. These informations are collected on a normal running system. The "graphics" file contains changes added to "/etc/system/enum/devices/graphics". After modifying "-d poulsbo" to "-d vga -d vesabios -d poulsbo" photon starts up but only in vesabios mode i.e. poulsbo driver does not run. The "graphlog.txt" is a system log capture of what happens when I try to change the graphic driver from vesabios to pulsbo. Regards, Maurizio Rossi

Maurizio Rossi 2008-10-15T13:01:41Z Keith Russell http://community.qnx.com/sf/go/post15007 2008-10-15T12:59:07Z 2008-10-15T12:59:07Z

I was reminded that this tech note was not present in the M8 build. The note and supporting files will be present in the next posted version. Sorry for the confusion. -- Keith

Keith Russell 2008-10-15T12:59:07Z Maurizio Rossi http://community.qnx.com/sf/go/post15005 2008-10-15T12:50:20Z 2008-10-15T12:50:20Z

Attached to this reply the results of some tests. Regards, Maurizio Rossi

Maurizio Rossi 2008-10-15T12:50:20Z Keith Russell http://community.qnx.com/sf/go/post15004 2008-10-15T12:47:35Z 2008-10-15T12:47:35Z

Yes, this is by design. The documentation would take up too much room uncompressed on a "live CD". The run from CD option is really to support the installation. It also allows some level of non-destructive testing to ensure QNX will run on your hardware. Cheers, Keith Russell -----Original Message----- From: Malte Mundt [mailto:community-noreply@qnx.com] Sent: October 15, 2008 08:22 To: sdp640prerelease-community Subject: No Docs / No Help When booting directly from CD into QNX Photon, and then open the Help Viewer, there is nothing in it. Is this by design? _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post14998

Keith Russell 2008-10-15T12:47:35Z Keith Russell http://community.qnx.com/sf/go/post15002 2008-10-15T12:43:49Z 2008-10-15T12:43:49Z

Malte, You are installing a development seat, not a runtime. The minimum required space is, I believe, around 1.6GB (but please consult the Installation Notes for confirmation). There is a tech note on how to create a runtime installation CD if all you want is the base OS. Cheers, Keith Russell -----Original Message----- From: Malte Mundt [mailto:community-noreply@qnx.com] Sent: October 15, 2008 08:21 To: sdp640prerelease-community Subject: Not enough room to install? Just tried to install QNX Neutrino on a Kontron nanoETXexpress from CD. I've seen this board working under QNX, however when trying to install, I get that there is not enough room to install. The board I have has a 512MB Flash EIDE disk. Shouldn't 512MB be enough for QNX to install? - Malte _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post14997

Keith Russell 2008-10-15T12:43:49Z Malte Mundt http://community.qnx.com/sf/go/post14998 2008-10-15T12:22:09Z 2008-10-15T12:22:09Z

When booting directly from CD into QNX Photon, and then open the Help Viewer, there is nothing in it. Is this by design?

Malte Mundt 2008-10-15T12:22:09Z Malte Mundt http://community.qnx.com/sf/go/post14997 2008-10-15T12:20:37Z 2008-10-15T12:20:37Z

Just tried to install QNX Neutrino on a Kontron nanoETXexpress from CD. I've seen this board working under QNX, however when trying to install, I get that there is not enough room to install. The board I have has a 512MB Flash EIDE disk. Shouldn't 512MB be enough for QNX to install? - Malte

Malte Mundt 2008-10-15T12:20:37Z Robert Craig http://community.qnx.com/sf/go/post14979 2008-10-14T18:17:08Z 2008-10-14T18:17:08Z

Hi Mathias: That log line is normal. QNET essentially looking through the possible interfaces and finding that no Ethernet device has been specified (which makes sense since you're using IP). This configuration now all works for me. I can get two VmWare sessions communicating with QNET over IP. Just to summarize : 1) Edit /etc/hosts on both machines (they should contain the same entries). Make sure that the domain as well as the host is set. (e.g. an entry like 192.168.23.128 node1.ott.qnx.com node1 192.168.23.129 node1.ott.qnx.com node2 ) 2) hostname node1 3) setconf _CS_DOMAIN ott.qnx.com 4) setconf _CS_RESOLVE lookup_file_bind 5) ping node1 6) ping node2 7) mount -Tio-pkt -o"bind=ip,resolve=dns" lsm-qnet.so (Likewise on remote machine). ls /net/node2 pidin -n node2 then work for me. One thing to remember for IP... the entries in /net aren't automatically populated. They're only populated by an attempt to access a remote node. Robert. -----Original Message----- From: Matthias Klein [mailto:community-noreply@qnx.com] Sent: Friday, October 10, 2008 5:51 PM To: sdp640prerelease-community Subject: Re: RE: RE: RE: RE: TDP over IP Hello Robert, > I suspect that the problem is that your domain doesn't match what's in /etc/hosts. Right. I changed /etc/hosts: old: 192.168.1.123 nas new: 192.168.1.123 nas.qnet nas Now I can "pidin -n nas" (local node), and gns also runs without errors. But I can't access the remote node. On both nodes the local access and the /net entry is OK. But in sloginfo of both node I get the following entry: Oct 10 23:39:51 7 15 0 qnet(L4): try_ifp(): ignoring interface en0 as per bind option What mean that entry ? Best regards Matthias _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post14890

Robert Craig 2008-10-14T18:17:08Z Robert Craig http://community.qnx.com/sf/go/post14963 2008-10-14T15:34:05Z 2008-10-14T15:34:05Z

Hi Maurizio: One of the things that came up with the devnp version of the driver is that we have some code in the devn driver which we don't know if we can put in the public domain. We ended up taking code from FreeBSD to do the EEPROM reading so it's possible that the devn driver may support more devices than the devnp (although I thought that we were the same). Could you tell me what the output from sloginfo reports when using the devnp driver? You'll also find that the devn driver (built for 6.4.0) is still included in the packaging for exactly this reason, so you can given that one a try as well. Thanks Robert. -----Original Message----- From: Maurizio Rossi [mailto:community-noreply@qnx.com] Sent: Monday, October 13, 2008 9:58 AM To: sdp640prerelease-community Subject: QNX6.4.0 M8 Build on Atom I'm testing 6.4.0 on a Kontron system based on Intel Atom processor. This board has the following devices: network adapter vid=0x8086 did=0x10d3 graphics adapter vid=0x8086 did=0x8108 I don't konw if what I'm reporting is a bug or not. The "devnp-i82544.so" is unable to manage the network controllers i.e. refuse to startup. The "devn-i82544.so" from QNX6.3.2 running on 6.4.0 works fine. In both cases the vid/did parameters are passed to the driver. Adding the "devg-poulsbo.so" driver to "/etc/system/enum/devices/graphics" cause the photon environment not to startup. Maybe this driver does not work with Photon? Regards, Maurizio Rossi _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post14911

Robert Craig 2008-10-14T15:34:05Z Aaron Cripps http://community.qnx.com/sf/go/post14950 2008-10-14T14:11:16Z 2008-10-14T14:11:16Z

Just as an experiment, restart your computer and after you select an image, hit the space bar for extra boot options, hit F6 two or three times and hit enter. Take note of the last message before booting halts. Also, on another startup, enable a debug shell, and when boot halts and kicks you to the shell, set your path and run "get_hw_info" and attach a copy of the results here (you can use a usb key to get the results onto a comp. with net access). I'm looking forward to hearing the results of these. Cheers, -Aaron On Monday 13 October 2008 09:58:03 Maurizio Rossi wrote: > I'm testing 6.4.0 on a Kontron system based on Intel Atom processor. > This board has the following devices: > network adapter vid=0x8086 did=0x10d3 > graphics adapter vid=0x8086 did=0x8108 > > I don't konw if what I'm reporting is a bug or not. > > The "devnp-i82544.so" is unable to manage the network controllers i.e. > refuse to startup. The "devn-i82544.so" from QNX6.3.2 running on 6.4.0 > works fine. > In both cases the vid/did parameters are passed to the driver. > > Adding the "devg-poulsbo.so" driver to "/etc/system/enum/devices/graphics" > cause the photon environment not to startup. Maybe this driver does not > work with Photon? > > Regards, > Maurizio Rossi > > > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post14911

Aaron Cripps 2008-10-14T14:11:16Z Derek Leach http://community.qnx.com/sf/go/post14948 2008-10-14T13:48:32Z 2008-10-14T13:48:32Z

to try

Derek Leach 2008-10-14T13:48:32Z Derek Leach http://community.qnx.com/sf/go/post14943 2008-10-14T13:18:11Z 2008-10-14T13:18:11Z

where did you get the devg-poulsbo from?

Derek Leach 2008-10-14T13:18:11Z Maurizio Rossi http://community.qnx.com/sf/go/post14911 2008-10-13T13:58:01Z 2008-10-13T13:58:01Z

I'm testing 6.4.0 on a Kontron system based on Intel Atom processor. This board has the following devices: network adapter vid=0x8086 did=0x10d3 graphics adapter vid=0x8086 did=0x8108 I don't konw if what I'm reporting is a bug or not. The "devnp-i82544.so" is unable to manage the network controllers i.e. refuse to startup. The "devn-i82544.so" from QNX6.3.2 running on 6.4.0 works fine. In both cases the vid/did parameters are passed to the driver. Adding the "devg-poulsbo.so" driver to "/etc/system/enum/devices/graphics" cause the photon environment not to startup. Maybe this driver does not work with Photon? Regards, Maurizio Rossi

Maurizio Rossi 2008-10-13T13:58:01Z Matthias Klein http://community.qnx.com/sf/go/post14890 2008-10-10T21:50:52Z 2008-10-10T21:50:52Z

Hello Robert, > I suspect that the problem is that your domain doesn't match what's in /etc/hosts. Right. I changed /etc/hosts: old: 192.168.1.123 nas new: 192.168.1.123 nas.qnet nas Now I can "pidin -n nas" (local node), and gns also runs without errors. But I can't access the remote node. On both nodes the local access and the /net entry is OK. But in sloginfo of both node I get the following entry: Oct 10 23:39:51 7 15 0 qnet(L4): try_ifp(): ignoring interface en0 as per bind option What mean that entry ? Best regards Matthias

Matthias Klein 2008-10-10T21:50:52Z Robert Craig http://community.qnx.com/sf/go/post14888 2008-10-10T21:12:37Z 2008-10-10T21:12:37Z

Hi Mathias: I think that I've figured it out. I suspect that the problem is that your domain doesn't match what's in /etc/hosts. setconf _CS_DOMAIN xxxxxxx 1) Make sure that your hostname is correctly configure using hostname 2) Make sure that your domain is correctly set (setconf _CS_DOMAIN ....) 3) Make sure your DNS resolver is correctly set (setconf _CS_RESOLVE ...) All of these steps have to be completed before you start qnet. If there's no local entry in /net, then qnet has failed. If any one of those three steps was missing, I'd end up with the same error conditions as you did. I've updated the Wiki as well. Robert. -----Original Message----- From: Matthias Klein [mailto:community-noreply@qnx.com] Sent: Friday, October 10, 2008 12:19 PM To: sdp640prerelease-community Subject: Re: RE: RE: RE: TDP over IP Hello Robert, > What command line are you using to start io-pkt? io-pkt-v4-hc -p tcpip forward -p qnet bind=ip,resolve=dns & waitfor /net enum-devices waitfor /dev/io-net/en0 ifconfig en0 192.168.1.123 up We used the forward option on QNX 6.3.2 to allow network access from modem dial-in via pppd. (Until now I didn't tested the modem dial-in on QNX 6.4) I also tested it without the forward option. The outputs of the command are in the attachment. Best regards Matthias _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post14864

Robert Craig 2008-10-10T21:12:37Z Matthias Klein http://community.qnx.com/sf/go/post14864 2008-10-10T16:18:57Z 2008-10-10T16:18:57Z

Hello Robert, > What command line are you using to start io-pkt? io-pkt-v4-hc -p tcpip forward -p qnet bind=ip,resolve=dns & waitfor /net enum-devices waitfor /dev/io-net/en0 ifconfig en0 192.168.1.123 up We used the forward option on QNX 6.3.2 to allow network access from modem dial-in via pppd. (Until now I didn't tested the modem dial-in on QNX 6.4) I also tested it without the forward option. The outputs of the command are in the attachment. Best regards Matthias

Matthias Klein 2008-10-10T16:18:57Z Robert Craig http://community.qnx.com/sf/go/post14862 2008-10-10T16:00:01Z 2008-10-10T16:00:01Z

Hi Mathias: Let's do a bit more debugging... If you aren't seeing a /net entry for the local machine then there's a fundamental problem showing up. What command line are you using to start io-pkt? Can you post the output from... pidin ar pidin -p io-pkt-v4-hc mem sloginfo Thanks! Robert. -----Original Message----- From: Matthias Klein [mailto:community-noreply@qnx.com] Sent: Friday, October 10, 2008 11:57 AM To: sdp640prerelease-community Subject: Re: RE: RE: TDP over IP Hello Robert, Setting the "_CS_RESOLVE" to "lookup_file_bind" doesn't change anything. Same errors ... > if you ping your hostname, does it resolve to the right IP address and work? Yes, the ping to the local hostname works. (Since I added the hostname to the /etc/hosts file) > Is there a /net entry created? Yes, there is a /net directory created. > Is it populated with a local entry? No, the /net directory is always empty. If it helps, I can upload the VMware 6 image (ca. 68 MB) to a webspace and send you a download link. Best regards Matthias _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post14861

Robert Craig 2008-10-10T16:00:01Z Matthias Klein http://community.qnx.com/sf/go/post14861 2008-10-10T15:57:00Z 2008-10-10T15:57:00Z

Hello Robert, Setting the "_CS_RESOLVE" to "lookup_file_bind" doesn't change anything. Same errors ... > if you ping your hostname, does it resolve to the right IP address and work? Yes, the ping to the local hostname works. (Since I added the hostname to the /etc/hosts file) > Is there a /net entry created? Yes, there is a /net directory created. > Is it populated with a local entry? No, the /net directory is always empty. If it helps, I can upload the VMware 6 image (ca. 68 MB) to a webspace and send you a download link. Best regards Matthias

Matthias Klein 2008-10-10T15:57:00Z Robert Craig http://community.qnx.com/sf/go/post14855 2008-10-10T15:38:15Z 2008-10-10T15:38:15Z

OK... I think I know what's going on. There is a mistake in the wiki about how to change the DNS lookup over to /etc/hosts. Changing resolv.conf doesn't do anything for you. You have to change the _CS_RESOLVE configuration string instead. setconf _CS_RESOLVE lookup_file_bind Let's work with the local case first... if you ping your hostname, does it resolve to the right IP address and work? Is there a /net entry created? Is it populated with a local entry? Robert -----Original Message----- From: Matthias Klein [mailto:community-noreply@qnx.com] Sent: Thursday, October 09, 2008 8:41 PM To: sdp640prerelease-community Subject: Re: RE: TDP over IP Hello Robert, >I still haven't had a chance to give this a shot (not surprisingly, >there's a lot going on these days). I did find this PR though. Does it >seem relevant? No, it seems not relevant. I can't get it work at all. No local node, no remote nodes. Even "gns -s" stops with the following error: netmgr_ndtostr(): No such process Best regards Matthias _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post14800

Robert Craig 2008-10-10T15:38:15Z Matthias Klein http://community.qnx.com/sf/go/post14800 2008-10-10T00:41:06Z 2008-10-10T00:41:06Z

Hello Robert, >I still haven't had a chance to give this a shot (not surprisingly, >there's a lot going on these days). I did find this PR though. Does it >seem relevant? No, it seems not relevant. I can't get it work at all. No local node, no remote nodes. Even "gns -s" stops with the following error: netmgr_ndtostr(): No such process Best regards Matthias

Matthias Klein 2008-10-10T00:41:06Z Robert Craig http://community.qnx.com/sf/go/post14798 2008-10-09T22:48:30Z 2008-10-09T22:48:30Z

I still haven't had a chance to give this a shot (not surprisingly, there's a lot going on these days). I did find this PR though. Does it seem relevant? When using the bind=ip,resolve=dns options, the resolver only name resolves the local machine name at startup. If qnet is started without name resolution working for the locally configured hostname, it will not work even if the name is added later. A re-start of qnet is required for the name to be resolved and entered in the database. Robert. -----Original Message----- From: Matthias Klein [mailto:community-noreply@qnx.com] Sent: Tuesday, October 07, 2008 2:00 PM To: sdp640prerelease-community Subject: TDP over IP Hello, I try to use TDP over IP on QNX 6.4 M8. I use the steps from this page: http://community.qnx.com/sf/wiki/do/viewPage/projects.networking/wiki/Qn et_wiki_page I can start io-pkt-v4-hc and TCP/IP works fine. But QNET didn't work at all. When I start e.g. pidin for the local node: pidin -n local_node_name I get the following output: pid tid name prio STATE Blocked pidin: could't find node local_node_name: Memory fault (I get the same error for a remote node) The /net directory is always empty. I use 2 VM's on VMware 6 for the two nodes. (TDP over ethernet works fine on the two "machines") Is the "Memory Fault" normal ? Any tips ? Best regards Matthias _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post14623

Robert Craig 2008-10-09T22:48:30Z Aleksandar Ristovski http://community.qnx.com/sf/go/post14792 2008-10-09T20:22:44Z 2008-10-09T20:22:44Z

What follows is not a supported scenario, it is something a few of us have been discussing as a "it would be nice if...". We are interested in your feedback since you brouhgt up the issue (we are also looking into fixing qconfig). Note that what follows is not officially supported; I am mostly interested in your feedback. Copy the attached qnxenv.cmd file into your QNX_HOST in both 6.3.2 and 6.4.0 installations. On my machine the destination directory is: C:\QNX632\host\win32\x86 and C:\QNX640\host\win32\x86 When you need to work with 6.3.2 environment, start "qnxenv.cmd" located in your 6.3.2 installation. It will open a command prompt (cmd.exe) with proper environment set, and from this cmd prompt you can start your shell or whatever is needed. Similarly, when you want to work with 6.4.0 environment, start "qnxenv.cmd" located in your 6.4.0 install dir and then start your make or shell from there. Your feedback would be greatly appreciated. Thanks

Aleksandar Ristovski 2008-10-09T20:22:44Z Alex Chapiro http://community.qnx.com/sf/go/post14739 2008-10-09T04:25:45Z 2008-10-09T04:25:45Z

AS I mentioned before, all conversion should occur automatically while you build it the first time. So no manual editing expected. If it does not happen it could be a bug. I'll check it once again when I'll be in the office. On 08/10/2008 8:01 PM, William Smith wrote: > We aren't using an IDE, but instead are using the makefile recurse mechanism for batch compiles. The "projects" that we have (loosely called that because they don't have the .project files) have worked with the 4.0 IDE. What comes with the 6.4 Alpha 8 is the 4.5 IDE. > > By creating a dummy project with the 4.5 IDE, I found if I replaced the "internal" section (like in the one I uploaded in the previous message) with the one generated by the 4.5 IDE, things worked better. Unfortunately that means editing a ton of common.mk files. > > _______________________________________________ > QNX Software Development Platform 6.4.0 Pre-Release > http://community.qnx.com/sf/go/post14733 > >

Alex Chapiro 2008-10-09T04:25:45Z William Smith http://community.qnx.com/sf/go/post14733 2008-10-09T00:01:27Z 2008-10-09T00:01:27Z

We aren't using an IDE, but instead are using the makefile recurse mechanism for batch compiles. The "projects" that we have (loosely called that because they don't have the .project files) have worked with the 4.0 IDE. What comes with the 6.4 Alpha 8 is the 4.5 IDE. By creating a dummy project with the 4.5 IDE, I found if I replaced the "internal" section (like in the one I uploaded in the previous message) with the one generated by the 4.5 IDE, things worked better. Unfortunately that means editing a ton of common.mk files.

William Smith 2008-10-09T00:01:27Z Derek Leach http://community.qnx.com/sf/go/post14727 2008-10-08T22:12:12Z 2008-10-08T22:12:12Z

we will post something for you asap

Derek Leach 2008-10-08T22:12:12Z Martin Gagnon http://community.qnx.com/sf/go/post14721 2008-10-08T20:48:24Z 2008-10-08T20:48:24Z

We are trying to make some test apps that integrate OpenGL drawing with Photon, however in Photon our apps don't work. egl-gears and other sample apps can work on layer 2 or 3, but we don't have the source for them. On various different video cards sometimes the gears will not show up in Photon, but still take resources and are running fine, still printing their framerate. If anyone could provide the source for the EGL demos or any source that integrates GF with photon we would like to see it. We want to create an application where OpenGL is drawn inside a Photon GUI window. We are using QNX 6.4 M8 alpha with Intel Extreme2 drivers (865 chipset). But also we get various different results using other intel chipsets or radeon cards. Thanks, NDT Team

Martin Gagnon 2008-10-08T20:48:24Z Alex Chapiro http://community.qnx.com/sf/go/post14715 2008-10-08T19:22:00Z 2008-10-08T19:22:00Z

Could you please let me know which IDE build you are using now? Which QNX SDK version? Some time ago we had a few bugs in this area, but they were fixed. AS for now all necessary conversion should occur behind the scene.

Alex Chapiro 2008-10-08T19:22:00Z William Smith http://community.qnx.com/sf/go/post14712 2008-10-08T17:02:43Z 2008-10-08T17:02:43Z

Some additional info. I was able to get it to work by using the 6.3 qconfig binary and then running the results through thru sed -e 's/\/cygdrive//g' -e 's|c:|/c|gi' While this works, I don't think this is the best solution.

William Smith 2008-10-08T17:02:43Z William Smith http://community.qnx.com/sf/go/post14710 2008-10-08T17:00:26Z 2008-10-08T17:00:26Z

I'm having a bit of trouble. We have an app (1 of many) that we have been compiling under 6.3.2. Attached is the common.mk that we have for it. Under 6.4.0, we get the following error when we get to the link phase: make[2]: *** No rule to make target `lib-Bstatic.a', needed by `/d/emb/qnx630/baseline/apps/src/actrld/ppc/o-be/actrld'. Stop. make[2]: Leaving directory `/d/emb/qnx630/baseline/apps/src/actrld/ppc/o-be' make[1]: *** [all] Error 2 make[1]: Leaving directory `/d/emb/qnx630/baseline/apps/src/actrld/ppc' Evidently there was some kind of change in the makefile structure. Is there any documentation on the makefile changes or how to get our 6.3.2 libraries and apps to compile? It only appears to be a problem when we want to statically link in a library. Thx, Bill

William Smith 2008-10-08T17:00:26Z Matthias Klein http://community.qnx.com/sf/go/post14691 2008-10-08T14:30:55Z 2008-10-08T14:30:55Z

Hello Frederic, can you add the VIA VT6105M chip (device ID 3053h) to the enum file ? Best regards Matthias

Matthias Klein 2008-10-08T14:30:55Z Frederic Plante http://community.qnx.com/sf/go/post14690 2008-10-08T14:03:41Z 2008-10-08T14:03:41Z

Hi Matthias, As Peter mentioned, we removed many drivers from 6.4 that were considered obsolete; for hardware that was no longer being sold. The devn-via-rhine driver was not part of the deprecated list and was mistakenly not included in the 6.4 ship list. Thanks to your feedback, the driver is back into 6.4. Got to love transparent development! :-) Thank you! :-) - Fred -----Original Message----- From: Matthias Klein [mailto:community-noreply@qnx.com] Sent: October 6, 2008 4:40 PM To: sdp640prerelease-community Subject: Re: RE: Missing: devn-via-rhine.so Hello Peter, > Many drivers for older hardware were not carried forward into 6.4 I can't understand why QNX decide that the VIA Rhine chipset is older hardware. Here are some current boards which use the VIA Rhine chipset: - Siemens SIMATIC Panel PC 677 - VIA EPIA PX-Series Pico-ITX Mainboards - VIA EPIA M-Series Mini-ITX Mainboards - ALIX mainboards from www.pcengines.ch ... Best regards Matthias _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post14546

Frederic Plante 2008-10-08T14:03:41Z Robert Craig http://community.qnx.com/sf/go/post14658 2008-10-08T02:02:41Z 2008-10-08T02:02:41Z

Hi Matthias: No... I'm afraid that you didn't miss anything. We're going through the final stages of the legal negotiations associated with allowing us to release the WiFi binaries to the public. Those are planned to be in the final release candidate coming up shortly. Robert. -----Original Message----- From: Matthias Klein [mailto:community-noreply@qnx.com] Sent: Tue 10/7/2008 9:15 PM To: sdp640prerelease-community Subject: Missing: WLAN drivers Is it right, that there are no WLAN drivers in QNX 6.4 M8 ? Do I miss something ? I want to test the new stack with a bcm43xx chip ... _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post14657

Robert Craig 2008-10-08T02:02:41Z Matthias Klein http://community.qnx.com/sf/go/post14657 2008-10-08T01:15:07Z 2008-10-08T01:15:07Z

Is it right, that there are no WLAN drivers in QNX 6.4 M8 ? Do I miss something ? I want to test the new stack with a bcm43xx chip ...

Matthias Klein 2008-10-08T01:15:07Z William Smith http://community.qnx.com/sf/go/post14656 2008-10-08T00:08:14Z 2008-10-08T00:08:14Z

We currently use a set of shell scripts for setting up our environment for different releases. To date, we have been using (per the docs) the command: eval `qconfig -n "QNX Momentics 6.3.0 SP3" -e`to set up the environment variables for different releases. For example, the output from the above qconfig command is: export QNX_HOST="C:/QNX630/host/win32/x86"; export QNX_TARGET="C:/QNX630/target/qnx6"; export PATH="/cygdrive/c/QNX630/host/win32/x86/usr/bin:/cygdrive/c/QNX630/host/win32/x86/bin:/cygdrive/c/QNX630/host/win32/x86/sbin:/cygdriv e/c/QNX630/host/win32/x86/usr/sbin:/cygdrive/c/QNX630/host/win32/x86/usr/photon/appbuilder:/cygdrive/c/Program Files/Mozilla Firefox:/cygdri ve/c/Perl/bin:/cygdrive/c/Python25/:/cygdrive/c/program files/imagemagick-6.3.0-q8:/cygdrive/c/Program Files/Java/jdk1.5.0_09/bin:/cygdrive/ c/PHP:/cygdrive/c/Perl/bin/:/cygdrive/d/bin:/cygdrive/c/WINDOWS/system32:/cygdrive/c/WINDOWS:/cygdrive/c/WINDOWS/System32/Wbem:/cygdrive/c/P rogram Files/Common Files/Roxio Shared/DLLShared:/cygdrive/c/program files/borland/StarTeam SDK 2005 R2/Lib:/cygdrive/c/program files/borlan d/StarTeam SDK 2005 R2/Bin:/cygdrive/c/Program Files/QNX Software Systems/bin:/cygdrive/c/Program Files/ATI Technologies/ATI Control Panel:/ cygdrive/c/Program Files/Common Files/GTK/2.0/bin:/cygdrive/c/Program Files/QuickTime/QTSystem/:/cygdrive/c/Program Files/TortoiseSVN/bin:/c ygdrive/c/Program Files/IDM Computer Solutions/UltraEdit/:/cygdrive/c/Program Files/IDM Computer Solutions/UltraCompare:/cygdrive/c/Program Files/IDM Computer Solutions/UltraCompare/:/cygdrive/c/Program Files/QNX Software Systems/bin:/cygdrive/c/Program Files/IDM Computer Solutio ns/UltraEdit/:/cygdrive/c/Program Files/IDM Computer Solutions/UltraCompare:/cygdrive/c/Program Files/IDM Computer Solutions/UltraCompare/"; export LD_LIBRARY_PATH="/cygdrive/c/QNX630/host/win32/x86/usr/lib"; export MAKEFLAGS="-IC:/QNX630/target/qnx6/usr/include"; This has been working fine. However today, when I installed 6.4.0 Alpha 8, the script I created like above failed. Here is the output from the 6.4.0 qconfig: export QNX_HOST="C:/QNX640/host/win32/x86"; export QNX_TARGET="C:/QNX640/target/qnx6"; export PATH="C:/QNX640/host/win32/x86/usr/bin:C:/QNX640/host/win32/x86/bin:C:/QNX640/host/win32/x86/sbin:C:/QNX640/host/win32/x86/usr/sbin:C :/QNX640/host/win32/x86/usr/photon/appbuilder:c:\Program Files\Mozilla Firefox;c:\Perl\bin;c:\Python25\;c:\program files\imagemagick-6.3.0-q 8;c:\Program Files\Java\jdk1.5.0_09\bin;c:\PHP;c:\Perl\bin\;d:\bin;c:\WINDOWS\system32;c:\WINDOWS;c:\WINDOWS\System32\Wbem;c:\Program Files\ Common Files\Roxio Shared\DLLShared;c:\program files\borland\StarTeam SDK 2005 R2\Lib;c:\program files\borland\StarTeam SDK 2005 R2\Bin;c:\P rogram Files\QNX Software Systems\bin;c:\Program Files\ATI Technologies\ATI Control Panel;c:\Program Files\Common Files\GTK\2.0\bin;c:\Progr am Files\QuickTime\QTSystem\;c:\Program Files\TortoiseSVN\bin;c:\Program Files\IDM Computer Solutions\UltraEdit\;c:\Program Files\IDM Comput er Solutions\UltraCompare;c:\Program Files\IDM Computer Solutions\UltraCompare\;c:\Program Files\QNX Software Systems\bin;c:\QNX630\host\win 32\x86\usr\bin;c:\Program Files\IDM Computer Solutions\UltraEdit\;c:\Program Files\IDM Computer Solutions\UltraCompare;c:\Program Files\IDM Computer Solutions\UltraCompare\"; export LD_LIBRARY_PATH="C:/QNX640/host/win32/x86/usr/lib"; export MAKEFLAGS="-IC:/QNX640/target/qnx6/usr/include"; Two things I noticed: 1) 6.4.0 appears to have dropped the /cygdrive/ part of the path in a shell. 2) The 6.4.0 qconfig leaves the values in "windows" form which blows up when you try and eval them in a shell script. Does anybody know a way to get the output from from the 6.4 qconfig to be in format that is functional in a bash shell script? BTW, if I use the 6.3 qconfig to select, it inserts the /cygdrive so they resulting variables are invalid. Thx, Bill

William Smith 2008-10-08T00:08:14Z Matthias Klein http://community.qnx.com/sf/go/post14623 2008-10-07T17:59:38Z 2008-10-07T17:59:38Z

Hello, I try to use TDP over IP on QNX 6.4 M8. I use the steps from this page: http://community.qnx.com/sf/wiki/do/viewPage/projects.networking/wiki/Qnet_wiki_page I can start io-pkt-v4-hc and TCP/IP works fine. But QNET didn't work at all. When I start e.g. pidin for the local node: pidin -n local_node_name I get the following output: pid tid name prio STATE Blocked pidin: could't find node local_node_name: Memory fault (I get the same error for a remote node) The /net directory is always empty. I use 2 VM's on VMware 6 for the two nodes. (TDP over ethernet works fine on the two "machines") Is the "Memory Fault" normal ? Any tips ? Best regards Matthias

Matthias Klein 2008-10-07T17:59:38Z Matthias Klein http://community.qnx.com/sf/go/post14546 2008-10-06T20:40:20Z 2008-10-06T20:40:20Z

Hello Peter, > Many drivers for older hardware were not carried forward into 6.4 I can't understand why QNX decide that the VIA Rhine chipset is older hardware. Here are some current boards which use the VIA Rhine chipset: - Siemens SIMATIC Panel PC 677 - VIA EPIA PX-Series Pico-ITX Mainboards - VIA EPIA M-Series Mini-ITX Mainboards - ALIX mainboards from www.pcengines.ch ... Best regards Matthias

Matthias Klein 2008-10-06T20:40:20Z Peter Manson http://community.qnx.com/sf/go/post14535 2008-10-06T16:59:53Z 2008-10-06T16:59:53Z

Matthias, Many drivers for older hardware were not carried forward into 6.4, and devn-via-rhine.so is one of them. This was decided some time ago, but the driver was just removed from the build recently (the docs still need to be removed). We have tested that the 6.3.2 binary driver works on a 6.4 target, so if you are using this device in 6.4, the 6.3.2 driver should work for you. Peter -----Original Message----- From: Matthias Klein [mailto:community-noreply@qnx.com] Sent: Sunday, October 5, 2008 9:43 PM To: sdp640prerelease-community Subject: Missing: devn-via-rhine.so Hello, what happened with devn-via-rhine.so in QNX 6.4 ? In the enum file of M7 the driver is listed. But in M7 + M8 is no binary file ... The help also lists the driver. Can you add the VIA VT6105M chip (device ID 3053h) to the enum file ? On QNX 6.3.2 the chip works fine with the devn-via-rhine.so driver. Best regards Matthias _______________________________________________ QNX Software Development Platform 6.4.0 Pre-Release http://community.qnx.com/sf/go/post14496

Peter Manson 2008-10-06T16:59:53Z