[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <40680C535D6FE6498883F1640FACD44DEA6714@ka-exchange-1.kontronamerica.local>
Date: Thu, 24 May 2012 13:05:05 -0700
From: "Andy Cress" <andy.cress@...kontron.com>
To: <netdev@...r.kernel.org>
Subject: pch_gbe: backport fails to start sending
Folks,
I now have a different case where the pch_gbe driver needs help.
I have backported the pch_gbe git head (v1.00) to kernel 2.6.32
(RHEL6.2) and when it loads, after the open completes and the link is
up, it fails to start sending and receiving.
I also took the pch_gbe source which runs fine on 2.6.38 (Fedora 15) and
backported it, and get the same results.
The much bulkier pch_gbe 0.91-NAPI driver does run on 2.6.32 (RHEL6.2)
but is not maintained.
I also tried backporting the mii.c from the 2.6.38 kernel, but that
didn't help, got the same symptoms.
After adding -DDEBUG to the 1.00 driver, I can see that it seems to get
just DMA Complete interrupts when it should be getting Transmit
complete, etc. I'm not sure why it gets stuck there.
Any ideas/input is welcome.
Andy
https://sendfile.kontron.com/message/KJKmrf171EhsuvbyhjnXpe
Attached at this link are two files:
pch_gbe-100a.tar.gz = the backported pch_gbe 1.00 head source, includes
patches that were applied.
dmesg.tar.gz = Some dmesg output from test cases with debug:
dmesg-pch10a-kern2632-bad.txt = backported 1.00 from git head on
kernel 2.6.32, fails
dmesg-pch10-kern2632-bad.txt = backported 1.00 from Fedora 15 on
kernel 2.6.32, fails
dmesg-pch10-kern2638-good.txt = same 1.00 source from Fedora 15 on
kernel 2.6.38, works
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists