gary anderson
|
mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx from 1 to 0
|
gary anderson
07/13/2010 12:22 PM
post59320
|
mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx from 1 to 0
I'm getting repetitive entries in the slog:
Jan 01 00:57:05 5 14 0 ient 80f9c3c from 0 to 1
Jan 01 00:57:05 5 14 0 ient 80f9c3c from 1 to 0
I get 15 or 16 entries per second. Any idea where they're coming from?
Thanks,
Gary
|
|
|
Hugh Brown
|
Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
from 1 to 0
|
Hugh Brown
07/13/2010 1:27 PM
post59323
|
Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
from 1 to 0
What network driver are you running? Are you using io-pkt or io-net?
On 10-07-13 12:22 PM, "gary anderson" <community-noreply@qnx.com> wrote:
> I'm getting repetitive entries in the slog:
>
> Jan 01 00:57:05 5 14 0 ient 80f9c3c from 0 to 1
> Jan 01 00:57:05 5 14 0 ient 80f9c3c from 1 to 0
>
> I get 15 or 16 entries per second. Any idea where they're coming from?
>
> Thanks,
> Gary
>
>
>
> _______________________________________________
>
> Networking Drivers
> http://community.qnx.com/sf/go/post59320
>
--
Hugh Brown (613) 591-0931 ext. 2209 (voice)
QNX Software Systems Ltd. (613) 591-3579 (fax)
175 Terence Matthews Cres. email: hsbrown@qnx.com
Kanata, Ontario, Canada.
K2M 1W8
|
|
|
gary anderson
|
RE: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
from 1 to 0
|
gary anderson
07/13/2010 1:31 PM
post59324
|
RE: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
from 1 to 0
io-pkt, QNX641.
-----Original Message-----
From: Hugh Brown [mailto:community-noreply@qnx.com]
Sent: Tuesday, July 13, 2010 10:28 AM
To: drivers-networking
Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx from 1 to 0
What network driver are you running? Are you using io-pkt or io-net?
On 10-07-13 12:22 PM, "gary anderson" <community-noreply@qnx.com> wrote:
> I'm getting repetitive entries in the slog:
>
> Jan 01 00:57:05 5 14 0 ient 80f9c3c from 0 to 1
> Jan 01 00:57:05 5 14 0 ient 80f9c3c from 1 to 0
>
> I get 15 or 16 entries per second. Any idea where they're coming from?
>
> Thanks,
> Gary
>
>
>
> _______________________________________________
>
> Networking Drivers
> http://community.qnx.com/sf/go/post59320
>
--
Hugh Brown (613) 591-0931 ext. 2209 (voice)
QNX Software Systems Ltd. (613) 591-3579 (fax)
175 Terence Matthews Cres. email: hsbrown@qnx.com
Kanata, Ontario, Canada.
K2M 1W8
_______________________________________________
Networking Drivers
http://community.qnx.com/sf/go/post59323
|
|
|
Hugh Brown
|
Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
from 1 to 0
|
Hugh Brown
07/13/2010 1:37 PM
post59326
|
Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
from 1 to 0
Yes, but what network driver are you running?
On 10-07-13 1:31 PM, "gary anderson" <community-noreply@qnx.com> wrote:
> io-pkt, QNX641.
>
>
> -----Original Message-----
> From: Hugh Brown [mailto:community-noreply@qnx.com]
> Sent: Tuesday, July 13, 2010 10:28 AM
> To: drivers-networking
> Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
> from 1 to 0
>
> What network driver are you running? Are you using io-pkt or io-net?
>
>
> On 10-07-13 12:22 PM, "gary anderson" <community-noreply@qnx.com> wrote:
>
>> I'm getting repetitive entries in the slog:
>>
>> Jan 01 00:57:05 5 14 0 ient 80f9c3c from 0 to 1
>> Jan 01 00:57:05 5 14 0 ient 80f9c3c from 1 to 0
>>
>> I get 15 or 16 entries per second. Any idea where they're coming from?
>>
>> Thanks,
>> Gary
>>
>>
>>
>> _______________________________________________
>>
>> Networking Drivers
>> http://community.qnx.com/sf/go/post59320
>>
--
Hugh Brown (613) 591-0931 ext. 2209 (voice)
QNX Software Systems Ltd. (613) 591-3579 (fax)
175 Terence Matthews Cres. email: hsbrown@qnx.com
Kanata, Ontario, Canada.
K2M 1W8
|
|
|
gary anderson
|
RE: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
from 1 to 0
|
gary anderson
07/13/2010 1:51 PM
post59328
|
RE: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
from 1 to 0
Running on an Intel platform with the devnp-e1000.so network driver plus another network driver (derived from the sam
ddk package) to implement to implement a private network over the PCIe bus. Jumbo packets are enabled (pagesize=8192,
mclbytes=8192) to improve performance of the PCIe implemented network.
-----Original Message-----
From: Hugh Brown [mailto:community-noreply@qnx.com]
Sent: Tuesday, July 13, 2010 10:37 AM
To: drivers-networking
Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx from 1 to 0
Yes, but what network driver are you running?
On 10-07-13 1:31 PM, "gary anderson" <community-noreply@qnx.com> wrote:
> io-pkt, QNX641.
>
>
> -----Original Message-----
> From: Hugh Brown [mailto:community-noreply@qnx.com]
> Sent: Tuesday, July 13, 2010 10:28 AM
> To: drivers-networking
> Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
> from 1 to 0
>
> What network driver are you running? Are you using io-pkt or io-net?
>
>
> On 10-07-13 12:22 PM, "gary anderson" <community-noreply@qnx.com> wrote:
>
>> I'm getting repetitive entries in the slog:
>>
>> Jan 01 00:57:05 5 14 0 ient 80f9c3c from 0 to 1
>> Jan 01 00:57:05 5 14 0 ient 80f9c3c from 1 to 0
>>
>> I get 15 or 16 entries per second. Any idea where they're coming from?
>>
>> Thanks,
>> Gary
>>
>>
>>
>> _______________________________________________
>>
>> Networking Drivers
>> http://community.qnx.com/sf/go/post59320
>>
--
Hugh Brown (613) 591-0931 ext. 2209 (voice)
QNX Software Systems Ltd. (613) 591-3579 (fax)
175 Terence Matthews Cres. email: hsbrown@qnx.com
Kanata, Ontario, Canada.
K2M 1W8
_______________________________________________
Networking Drivers
http://community.qnx.com/sf/go/post59326
|
|
|
Hugh Brown
|
Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
from 1 to 0
|
Hugh Brown
07/13/2010 2:00 PM
post59332
|
Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
from 1 to 0
There is no message like this in the e1000 driver, so I guess it could be
somewhere in the stack. ient is an interrupt entry, so do you have an
interrupt service routine in your driver?
On 10-07-13 1:51 PM, "gary anderson" <community-noreply@qnx.com> wrote:
> Running on an Intel platform with the devnp-e1000.so network driver plus
> another network driver (derived from the sam ddk package) to implement to
> implement a private network over the PCIe bus. Jumbo packets are enabled
> (pagesize=8192,mclbytes=8192) to improve performance of the PCIe implemented
> network.
>
> -----Original Message-----
> From: Hugh Brown [mailto:community-noreply@qnx.com]
> Sent: Tuesday, July 13, 2010 10:37 AM
> To: drivers-networking
> Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
> from 1 to 0
>
> Yes, but what network driver are you running?
>
>
> On 10-07-13 1:31 PM, "gary anderson" <community-noreply@qnx.com> wrote:
>
>> io-pkt, QNX641.
>>
>>
>> -----Original Message-----
>> From: Hugh Brown [mailto:community-noreply@qnx.com]
>> Sent: Tuesday, July 13, 2010 10:28 AM
>> To: drivers-networking
>> Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
>> from 1 to 0
>>
>> What network driver are you running? Are you using io-pkt or io-net?
>>
>>
>> On 10-07-13 12:22 PM, "gary anderson" <community-noreply@qnx.com> wrote:
>>
>>> I'm getting repetitive entries in the slog:
>>>
>>> Jan 01 00:57:05 5 14 0 ient 80f9c3c from 0 to 1
>>> Jan 01 00:57:05 5 14 0 ient 80f9c3c from 1 to 0
>>>
>>> I get 15 or 16 entries per second. Any idea where they're coming from?
>>>
>>> Thanks,
>>> Gary
>>>
>>>
>>>
>>> _______________________________________________
>>>
>>> Networking Drivers
>>> http://community.qnx.com/sf/go/post59320
>>>
--
Hugh Brown (613) 591-0931 ext. 2209 (voice)
QNX Software Systems Ltd. (613) 591-3579 (fax)
175 Terence Matthews Cres. email: hsbrown@qnx.com
Kanata, Ontario, Canada.
K2M 1W8
|
|
|
Sean Boudreau(deleted)
|
Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient
xxxxxxx from 1 to 0
|
Sean Boudreau(deleted)
07/13/2010 2:03 PM
post59333
|
Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient
xxxxxxx from 1 to 0
Yes, from the stack.
-seanb
On Tue, Jul 13, 2010 at 02:00:18PM -0400, Hugh Brown wrote:
> There is no message like this in the e1000 driver, so I guess it could be
> somewhere in the stack. ient is an interrupt entry, so do you have an
> interrupt service routine in your driver?
>
>
> On 10-07-13 1:51 PM, "gary anderson" <community-noreply@qnx.com> wrote:
>
> > Running on an Intel platform with the devnp-e1000.so network driver plus
> > another network driver (derived from the sam ddk package) to implement to
> > implement a private network over the PCIe bus. Jumbo packets are enabled
> > (pagesize=8192,mclbytes=8192) to improve performance of the PCIe implemented
> > network.
> >
> > -----Original Message-----
> > From: Hugh Brown [mailto:community-noreply@qnx.com]
> > Sent: Tuesday, July 13, 2010 10:37 AM
> > To: drivers-networking
> > Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
> > from 1 to 0
> >
> > Yes, but what network driver are you running?
> >
> >
> > On 10-07-13 1:31 PM, "gary anderson" <community-noreply@qnx.com> wrote:
> >
> >> io-pkt, QNX641.
> >>
> >>
> >> -----Original Message-----
> >> From: Hugh Brown [mailto:community-noreply@qnx.com]
> >> Sent: Tuesday, July 13, 2010 10:28 AM
> >> To: drivers-networking
> >> Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
> >> from 1 to 0
> >>
> >> What network driver are you running? Are you using io-pkt or io-net?
> >>
> >>
> >> On 10-07-13 12:22 PM, "gary anderson" <community-noreply@qnx.com> wrote:
> >>
> >>> I'm getting repetitive entries in the slog:
> >>>
> >>> Jan 01 00:57:05 5 14 0 ient 80f9c3c from 0 to 1
> >>> Jan 01 00:57:05 5 14 0 ient 80f9c3c from 1 to 0
> >>>
> >>> I get 15 or 16 entries per second. Any idea where they're coming from?
> >>>
> >>> Thanks,
> >>> Gary
> >>>
> >>>
> >>>
> >>> _______________________________________________
> >>>
> >>> Networking Drivers
> >>> http://community.qnx.com/sf/go/post59320
> >>>
>
> --
> Hugh Brown (613) 591-0931 ext. 2209 (voice)
> QNX Software Systems Ltd. (613) 591-3579 (fax)
> 175 Terence Matthews Cres. email: hsbrown@qnx.com
> Kanata, Ontario, Canada.
> K2M 1W8
>
>
>
>
>
>
> _______________________________________________
>
> Networking Drivers
> http://community.qnx.com/sf/go/post59332
>
|
|
|
gary anderson
|
RE: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient
xxxxxxx from 1 to 0
|
gary anderson
07/13/2010 6:08 PM
post59357
|
RE: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient
xxxxxxx from 1 to 0
I have a multiple CPU system with a PCIe network connection between the Monarch processor and every Endpoint processor.
In my startup code, as I discover CPU modules I execute a mount command to instantiate a driver for each discovered CPU
-- "mount -Tio-pkt -o<slot> libpcielan.so" where <slot> is passed to the driver to inform it which CPU to connect to
and it's also used to generate IP addresses. In my test system I have four CPUs -- the Monarch and three Endpoints so
three "mounts" are executed. After everything is init'ed I have the following ifconfig output from the Monarch CPU:
# ifconfig
lo0: flags=8049<UP,LOOPBACK,RUNNING,MULTICAST> mtu 33192
inet 127.0.0.1 netmask 0xff000000
wm0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
capabilities rx=1f<IP4CSUM,TCP4CSUM,UDP4CSUM,TCP6CSUM,UDP6CSUM>
capabilities tx=7f<IP4CSUM,TCP4CSUM,UDP4CSUM,TCP6CSUM,UDP6CSUM,TSO4,TSO6>
enabled=0
address: 00:17:3c:00:9a:60
media: Ethernet autoselect (1000baseT full-duplex)
status: active
inet 141.121.191.225 netmask 0xffffff00 broadcast 141.121.191.255
PCIe0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
address: 00:01:02:03:00:07
inet 10.1.7.7 netmask 0xffffff00 broadcast 10.1.7.255
PCIe1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
address: 00:01:02:03:00:05
inet 10.1.5.5 netmask 0xffffff00 broadcast 10.1.5.255
PCIe2: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
address: 00:01:02:03:00:0b
inet 10.1.11.11 netmask 0xffffff00 broadcast 10.1.11.255
At this point I start getting the "ient xxxxxxxx from 0 to 1, ient xxxxxxx from 1 to 0" entries in the syslog. If I
disable one of the PCIe NICs (ifconfig PCIe2 down) the syslog entries quit. Re-enabling the PCIe NIC causes the syslog
entries to resume.
The Monarch processor is configured with NAT to get from the public wm0 (Ethernet) port to the private network of
Endpoint CPUs. It seemingly all works but I don't understand the syslog entries.
Any ideas?
-----Original Message-----
From: Sean Boudreau [mailto:community-noreply@qnx.com]
Sent: Tuesday, July 13, 2010 11:04 AM
To: drivers-networking
Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx from 1 to 0
Yes, from the stack.
-seanb
On Tue, Jul 13, 2010 at 02:00:18PM -0400, Hugh Brown wrote:
> There is no message like this in the e1000 driver, so I guess it could be
> somewhere in the stack. ient is an interrupt entry, so do you have an
> interrupt service routine in your driver?
>
>
> On 10-07-13 1:51 PM, "gary anderson" <community-noreply@qnx.com> wrote:
>
> > Running on an Intel platform with the devnp-e1000.so network driver plus
> > another network driver (derived from the sam ddk package) to implement to
> > implement a private network over the PCIe bus. Jumbo packets are enabled
> > (pagesize=8192,mclbytes=8192) to improve performance of the PCIe implemented
> > network.
> >
> > -----Original Message-----
> > From: Hugh Brown [mailto:community-noreply@qnx.com]
> > Sent: Tuesday, July 13, 2010 10:37 AM
> > To: drivers-networking
> > Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
> > from 1 to 0
> >
> > Yes, but what network driver are you running?
> >
> >
> > On 10-07-13 1:31 PM, "gary anderson" <community-noreply@qnx.com> wrote:
> >
> >> io-pkt, QNX641.
> >>
> >>
> >> -----Original Message-----
> >> From: Hugh Brown [mailto:community-noreply@qnx.com]
> >> Sent: Tuesday, July 13, 2010 10:28 AM
> >> To: drivers-networking
> >> Subject: Re: mystery slog entries --...
View Full Message
|
|
|
Sean Boudreau(deleted)
|
Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient
xxxxxxx from 1 to 0
|
Sean Boudreau(deleted)
07/14/2010 9:19 AM
post59402
|
Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient
xxxxxxx from 1 to 0
On Tue, Jul 13, 2010 at 06:08:16PM -0400, gary anderson wrote:
> I have a multiple CPU system with a PCIe network connection between the Monarch processor and every Endpoint processor
. In my startup code, as I discover CPU modules I execute a mount command to instantiate a driver for each discovered
CPU -- "mount -Tio-pkt -o<slot> libpcielan.so" where <slot> is passed to the driver to inform it which CPU to connect
to and it's also used to generate IP addresses. In my test system I have four CPUs -- the Monarch and three Endpoints
so three "mounts" are executed. After everything is init'ed I have the following ifconfig output from the Monarch CPU:
>
> # ifconfig
> lo0: flags=8049<UP,LOOPBACK,RUNNING,MULTICAST> mtu 33192
> inet 127.0.0.1 netmask 0xff000000
> wm0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
> capabilities rx=1f<IP4CSUM,TCP4CSUM,UDP4CSUM,TCP6CSUM,UDP6CSUM>
> capabilities tx=7f<IP4CSUM,TCP4CSUM,UDP4CSUM,TCP6CSUM,UDP6CSUM,TSO4,TSO6>
> enabled=0
> address: 00:17:3c:00:9a:60
> media: Ethernet autoselect (1000baseT full-duplex)
> status: active
> inet 141.121.191.225 netmask 0xffffff00 broadcast 141.121.191.255
> PCIe0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
> address: 00:01:02:03:00:07
> inet 10.1.7.7 netmask 0xffffff00 broadcast 10.1.7.255
> PCIe1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
> address: 00:01:02:03:00:05
> inet 10.1.5.5 netmask 0xffffff00 broadcast 10.1.5.255
> PCIe2: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
> address: 00:01:02:03:00:0b
> inet 10.1.11.11 netmask 0xffffff00 broadcast 10.1.11.255
>
> At this point I start getting the "ient xxxxxxxx from 0 to 1, ient xxxxxxx from 1 to 0" entries in the syslog. If I
disable one of the PCIe NICs (ifconfig PCIe2 down) the syslog entries quit. Re-enabling the PCIe NIC causes the syslog
entries to resume.
>
> The Monarch processor is configured with NAT to get from the public wm0 (Ethernet) port to the private network of
Endpoint CPUs. It seemingly all works but I don't understand the syslog entries.
>
> Any ideas?
It's a debug slog from the stack. It's safe to ignore.
Regards,
-seanb
|
|
|
gary anderson
|
RE: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient
xxxxxxx from 1 to 0
|
gary anderson
07/14/2010 10:31 AM
post59419
|
RE: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient
xxxxxxx from 1 to 0
Is there any way to disable it? It is persistent and fills the syslog making the syslog unusable for any other logging.
Thanks,
Gary
-----Original Message-----
From: Sean Boudreau [mailto:community-noreply@qnx.com]
Sent: Wednesday, July 14, 2010 6:19 AM
To: drivers-networking
Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx from 1 to 0
On Tue, Jul 13, 2010 at 06:08:16PM -0400, gary anderson wrote:
> I have a multiple CPU system with a PCIe network connection between the Monarch processor and every Endpoint processor
. In my startup code, as I discover CPU modules I execute a mount command to instantiate a driver for each discovered
CPU -- "mount -Tio-pkt -o<slot> libpcielan.so" where <slot> is passed to the driver to inform it which CPU to connect
to and it's also used to generate IP addresses. In my test system I have four CPUs -- the Monarch and three Endpoints
so three "mounts" are executed. After everything is init'ed I have the following ifconfig output from the Monarch CPU:
>
> # ifconfig
> lo0: flags=8049<UP,LOOPBACK,RUNNING,MULTICAST> mtu 33192
> inet 127.0.0.1 netmask 0xff000000
> wm0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
> capabilities rx=1f<IP4CSUM,TCP4CSUM,UDP4CSUM,TCP6CSUM,UDP6CSUM>
> capabilities tx=7f<IP4CSUM,TCP4CSUM,UDP4CSUM,TCP6CSUM,UDP6CSUM,TSO4,TSO6>
> enabled=0
> address: 00:17:3c:00:9a:60
> media: Ethernet autoselect (1000baseT full-duplex)
> status: active
> inet 141.121.191.225 netmask 0xffffff00 broadcast 141.121.191.255
> PCIe0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
> address: 00:01:02:03:00:07
> inet 10.1.7.7 netmask 0xffffff00 broadcast 10.1.7.255
> PCIe1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
> address: 00:01:02:03:00:05
> inet 10.1.5.5 netmask 0xffffff00 broadcast 10.1.5.255
> PCIe2: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
> address: 00:01:02:03:00:0b
> inet 10.1.11.11 netmask 0xffffff00 broadcast 10.1.11.255
>
> At this point I start getting the "ient xxxxxxxx from 0 to 1, ient xxxxxxx from 1 to 0" entries in the syslog. If I
disable one of the PCIe NICs (ifconfig PCIe2 down) the syslog entries quit. Re-enabling the PCIe NIC causes the syslog
entries to resume.
>
> The Monarch processor is configured with NAT to get from the public wm0 (Ethernet) port to the private network of
Endpoint CPUs. It seemingly all works but I don't understand the syslog entries.
>
> Any ideas?
It's a debug slog from the stack. It's safe to ignore.
Regards,
-seanb
_______________________________________________
Networking Drivers
http://community.qnx.com/sf/go/post59402
|
|
|
Sean Boudreau(deleted)
|
Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient
xxxxxxx from 1 to 0
|
Sean Boudreau(deleted)
07/14/2010 10:34 AM
post59422
|
Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient
xxxxxxx from 1 to 0
Not currently. A workaround may be 'io-pkt -t4 ...' where 4
is your number of interfaces. Or 'sloginfo | grep -v ient'.
I'll make a PR to disable it or only spit it out with
io-pkt -v.
-seanb
On Wed, Jul 14, 2010 at 10:31:12AM -0400, gary anderson wrote:
> Is there any way to disable it? It is persistent and fills the syslog making the syslog unusable for any other
logging.
>
> Thanks,
> Gary
>
> -----Original Message-----
> From: Sean Boudreau [mailto:community-noreply@qnx.com]
> Sent: Wednesday, July 14, 2010 6:19 AM
> To: drivers-networking
> Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx from 1 to 0
>
> On Tue, Jul 13, 2010 at 06:08:16PM -0400, gary anderson wrote:
> > I have a multiple CPU system with a PCIe network connection between the Monarch processor and every Endpoint
processor. In my startup code, as I discover CPU modules I execute a mount command to instantiate a driver for each
discovered CPU -- "mount -Tio-pkt -o<slot> libpcielan.so" where <slot> is passed to the driver to inform it which CPU
to connect to and it's also used to generate IP addresses. In my test system I have four CPUs -- the Monarch and three
Endpoints so three "mounts" are executed. After everything is init'ed I have the following ifconfig output from the
Monarch CPU:
> >
> > # ifconfig
> > lo0: flags=8049<UP,LOOPBACK,RUNNING,MULTICAST> mtu 33192
> > inet 127.0.0.1 netmask 0xff000000
> > wm0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
> > capabilities rx=1f<IP4CSUM,TCP4CSUM,UDP4CSUM,TCP6CSUM,UDP6CSUM>
> > capabilities tx=7f<IP4CSUM,TCP4CSUM,UDP4CSUM,TCP6CSUM,UDP6CSUM,TSO4,TSO6>
> > enabled=0
> > address: 00:17:3c:00:9a:60
> > media: Ethernet autoselect (1000baseT full-duplex)
> > status: active
> > inet 141.121.191.225 netmask 0xffffff00 broadcast 141.121.191.255
> > PCIe0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
> > address: 00:01:02:03:00:07
> > inet 10.1.7.7 netmask 0xffffff00 broadcast 10.1.7.255
> > PCIe1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
> > address: 00:01:02:03:00:05
> > inet 10.1.5.5 netmask 0xffffff00 broadcast 10.1.5.255
> > PCIe2: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
> > address: 00:01:02:03:00:0b
> > inet 10.1.11.11 netmask 0xffffff00 broadcast 10.1.11.255
> >
> > At this point I start getting the "ient xxxxxxxx from 0 to 1, ient xxxxxxx from 1 to 0" entries in the syslog. If I
disable one of the PCIe NICs (ifconfig PCIe2 down) the syslog entries quit. Re-enabling the PCIe NIC causes the syslog
entries to resume.
> >
> > The Monarch processor is configured with NAT to get from the public wm0 (Ethernet) port to the private network of
Endpoint CPUs. It seemingly all works but I don't understand the syslog entries.
> >
> > Any ideas?
>
> It's a debug slog from the stack. It's safe to ignore.
>
> Regards,
>
> -seanb
>
>
>
> _______________________________________________
>
> Networking Drivers
> http://community.qnx.com/sf/go/post59402
>
>
>
>
> _______________________________________________
>
> Networking Drivers
> http://community.qnx.com/sf/go/post59419
>
|
|
|
Vineet Garg
|
Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
from 1 to 0
|
Vineet Garg
07/14/2010 12:13 PM
post59442
|
Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
from 1 to 0
Hi Sean
I have asked this question earlier but since you mention about debug traces
of io-pkt here and enabling them thru io-pkt -v
Can you please let me know if there is a way to get io-net debug traces in
systlog? I have tried doing io-net -v but that doesn't seem to help.
Regards
Vineet
On Wed, Jul 14, 2010 at 8:04 PM, Sean Boudreau <community-noreply@qnx.com>wrote:
>
>
> Not currently. A workaround may be 'io-pkt -t4 ...' where 4
> is your number of interfaces. Or 'sloginfo | grep -v ient'.
> I'll make a PR to disable it or only spit it out with
> io-pkt -v.
>
> -seanb
>
> On Wed, Jul 14, 2010 at 10:31:12AM -0400, gary anderson wrote:
> > Is there any way to disable it? It is persistent and fills the syslog
> making the syslog unusable for any other logging.
> >
> > Thanks,
> > Gary
> >
> > -----Original Message-----
> > From: Sean Boudreau [mailto:community-noreply@qnx.com]
> > Sent: Wednesday, July 14, 2010 6:19 AM
> > To: drivers-networking
> > Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient
> xxxxxxx from 1 to 0
> >
> > On Tue, Jul 13, 2010 at 06:08:16PM -0400, gary anderson wrote:
> > > I have a multiple CPU system with a PCIe network connection between the
> Monarch processor and every Endpoint processor. In my startup code, as I
> discover CPU modules I execute a mount command to instantiate a driver for
> each discovered CPU -- "mount -Tio-pkt -o<slot> libpcielan.so" where <slot>
> is passed to the driver to inform it which CPU to connect to and it's also
> used to generate IP addresses. In my test system I have four CPUs -- the
> Monarch and three Endpoints so three "mounts" are executed. After
> everything is init'ed I have the following ifconfig output from the Monarch
> CPU:
> > >
> > > # ifconfig
> > > lo0: flags=8049<UP,LOOPBACK,RUNNING,MULTICAST> mtu 33192
> > > inet 127.0.0.1 netmask 0xff000000
> > > wm0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
> > > capabilities rx=1f<IP4CSUM,TCP4CSUM,UDP4CSUM,TCP6CSUM,UDP6CSUM>
> > > capabilities
> tx=7f<IP4CSUM,TCP4CSUM,UDP4CSUM,TCP6CSUM,UDP6CSUM,TSO4,TSO6>
> > > enabled=0
> > > address: 00:17:3c:00:9a:60
> > > media: Ethernet autoselect (1000baseT full-duplex)
> > > status: active
> > > inet 141.121.191.225 netmask 0xffffff00 broadcast
> 141.121.191.255
> > > PCIe0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
> > > address: 00:01:02:03:00:07
> > > inet 10.1.7.7 netmask 0xffffff00 broadcast 10.1.7.255
> > > PCIe1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
> > > address: 00:01:02:03:00:05
> > > inet 10.1.5.5 netmask 0xffffff00 broadcast 10.1.5.255
> > > PCIe2: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 8100
> > > address: 00:01:02:03:00:0b
> > > inet 10.1.11.11 netmask 0xffffff00 broadcast 10.1.11.255
> > >
> > > At this point I start getting the "ient xxxxxxxx from 0 to 1, ient
> xxxxxxx from 1 to 0" entries in the syslog. If I disable one of the PCIe
> NICs (ifconfig PCIe2 down) the syslog entries quit. Re-enabling the PCIe NIC
> causes the syslog entries to resume.
> > >
> > > The Monarch processor is configured with NAT to get from the public wm0
> (Ethernet) port to the private network of Endpoint CPUs. It seemingly all
> works but I don't understand the syslog entries.
> > >
> > > Any ideas?
> >
> > It's a...
View Full Message
|
|
|
gary anderson
|
RE: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
from 1 to 0
|
gary anderson
07/13/2010 2:06 PM
post59334
|
RE: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
from 1 to 0
Right now my PCIe network driver is run off the timer interrupt (as done in the 'sam' example).
-----Original Message-----
From: Hugh Brown [mailto:community-noreply@qnx.com]
Sent: Tuesday, July 13, 2010 11:00 AM
To: drivers-networking
Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx from 1 to 0
There is no message like this in the e1000 driver, so I guess it could be
somewhere in the stack. ient is an interrupt entry, so do you have an
interrupt service routine in your driver?
On 10-07-13 1:51 PM, "gary anderson" <community-noreply@qnx.com> wrote:
> Running on an Intel platform with the devnp-e1000.so network driver plus
> another network driver (derived from the sam ddk package) to implement to
> implement a private network over the PCIe bus. Jumbo packets are enabled
> (pagesize=8192,mclbytes=8192) to improve performance of the PCIe implemented
> network.
>
> -----Original Message-----
> From: Hugh Brown [mailto:community-noreply@qnx.com]
> Sent: Tuesday, July 13, 2010 10:37 AM
> To: drivers-networking
> Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
> from 1 to 0
>
> Yes, but what network driver are you running?
>
>
> On 10-07-13 1:31 PM, "gary anderson" <community-noreply@qnx.com> wrote:
>
>> io-pkt, QNX641.
>>
>>
>> -----Original Message-----
>> From: Hugh Brown [mailto:community-noreply@qnx.com]
>> Sent: Tuesday, July 13, 2010 10:28 AM
>> To: drivers-networking
>> Subject: Re: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx
>> from 1 to 0
>>
>> What network driver are you running? Are you using io-pkt or io-net?
>>
>>
>> On 10-07-13 12:22 PM, "gary anderson" <community-noreply@qnx.com> wrote:
>>
>>> I'm getting repetitive entries in the slog:
>>>
>>> Jan 01 00:57:05 5 14 0 ient 80f9c3c from 0 to 1
>>> Jan 01 00:57:05 5 14 0 ient 80f9c3c from 1 to 0
>>>
>>> I get 15 or 16 entries per second. Any idea where they're coming from?
>>>
>>> Thanks,
>>> Gary
>>>
>>>
>>>
>>> _______________________________________________
>>>
>>> Networking Drivers
>>> http://community.qnx.com/sf/go/post59320
>>>
--
Hugh Brown (613) 591-0931 ext. 2209 (voice)
QNX Software Systems Ltd. (613) 591-3579 (fax)
175 Terence Matthews Cres. email: hsbrown@qnx.com
Kanata, Ontario, Canada.
K2M 1W8
_______________________________________________
Networking Drivers
http://community.qnx.com/sf/go/post59332
|
|
|
Ken Beoughter(deleted)
|
Re: RE: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx from 1 to 0
|
Ken Beoughter(deleted)
08/05/2010 2:46 PM
post61850
|
Re: RE: mystery slog entries -- ient xxxxxxxx from 0 to 1, ient xxxxxxx from 1 to 0
Is there anyway I can get the source to the PCIe lan implementation that Gary did?
|
|
|
|