* Re: [PATCH] w35und: fix usb_control_msg() error handling in wb35_probe() [not found] ` <497E0F94.8000003@gmail.com> @ 2009-01-26 19:40 ` Pekka Enberg 2009-01-26 19:43 ` Pekka Enberg 0 siblings, 1 reply; 4+ messages in thread From: Pekka Enberg @ 2009-01-26 19:40 UTC (permalink / raw) To: Sandro Bonazzola Cc: Pavel Machek, Greg Kroah-Hartman, linux-kernel, Linux Netdev List Hi Sandro, Pekka Enberg ha scritto: >> Ok, test done with git current version of w35und over 2.6.28-gentoo-r1 >> (2.6.28.1 and some patches) > > Adding Pekka's patch On Mon, Jan 26, 2009 at 9:31 PM, Sandro Bonazzola <sandro.bonazzola@gmail•com> wrote: >> # uname -a >> Linux arilinn 2.6.28-gentoo-r1 #1 Wed Jan 21 19:37:07 CET 2009 x86_64 AMD >> Athlon(tm) 64 Processor 3500+ AuthenticAMD GNU/Linux [snip] > Ok, with kernel 2.6.8.1, and your patch it seems to work with channel 1, no > encryption, opensystem authentication. So the driver doesn't work in 2.6.29-rc1 but does in 2.6.28.1? I wonder what changed in the networking/wireless stack that broke the winbond driver.... I'm cc'ing netdev even though this is a driver in staging. Pekka ^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH] w35und: fix usb_control_msg() error handling in wb35_probe() 2009-01-26 19:40 ` [PATCH] w35und: fix usb_control_msg() error handling in wb35_probe() Pekka Enberg @ 2009-01-26 19:43 ` Pekka Enberg 2009-01-26 20:13 ` Sandro Bonazzola 0 siblings, 1 reply; 4+ messages in thread From: Pekka Enberg @ 2009-01-26 19:43 UTC (permalink / raw) To: Sandro Bonazzola Cc: Pavel Machek, Greg Kroah-Hartman, linux-kernel, Linux Netdev List On Mon, Jan 26, 2009 at 9:40 PM, Pekka Enberg <penberg@cs•helsinki.fi> wrote: > Hi Sandro, > > Pekka Enberg ha scritto: >>> Ok, test done with git current version of w35und over 2.6.28-gentoo-r1 >>> (2.6.28.1 and some patches) >> >> Adding Pekka's patch > > On Mon, Jan 26, 2009 at 9:31 PM, Sandro Bonazzola > <sandro.bonazzola@gmail•com> wrote: >>> # uname -a >>> Linux arilinn 2.6.28-gentoo-r1 #1 Wed Jan 21 19:37:07 CET 2009 x86_64 AMD >>> Athlon(tm) 64 Processor 3500+ AuthenticAMD GNU/Linux > > [snip] > >> Ok, with kernel 2.6.8.1, and your patch it seems to work with channel 1, no >> encryption, opensystem authentication. > > So the driver doesn't work in 2.6.29-rc1 but does in 2.6.28.1? I > wonder what changed in the networking/wireless stack that broke the > winbond driver.... I'm cc'ing netdev even though this is a driver in > staging. Actually, it's probably me who broke the driver. I think Greg merged most of my patches after 2.6.28 came out. Sandro, if you have some time, please consider doing git bisect to find out which changeset causes the problem. Pekka ^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH] w35und: fix usb_control_msg() error handling in wb35_probe() 2009-01-26 19:43 ` Pekka Enberg @ 2009-01-26 20:13 ` Sandro Bonazzola 2009-01-26 20:22 ` Pekka Enberg 0 siblings, 1 reply; 4+ messages in thread From: Sandro Bonazzola @ 2009-01-26 20:13 UTC (permalink / raw) To: Pekka Enberg Cc: Pavel Machek, Greg Kroah-Hartman, linux-kernel, Linux Netdev List -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Pekka Enberg ha scritto: >>> Ok, with kernel 2.6.8.1, and your patch it seems to work with channel 1, no >>> encryption, opensystem authentication. >> So the driver doesn't work in 2.6.29-rc1 but does in 2.6.28.1? I >> wonder what changed in the networking/wireless stack that broke the >> winbond driver.... I'm cc'ing netdev even though this is a driver in >> staging. Well, I don't know if this can be somehow related to a bug in the driver, but on my system, after I've unplugged the USB trasnsitter, ksoftirqd/0 begin to use 89% of my CPU time (top). I suspect that there are some issue that should be fixed somewhere. > Actually, it's probably me who broke the driver. I think Greg merged > most of my patches after 2.6.28 came out. Sandro, if you have some > time, please consider doing git bisect to find out which changeset > causes the problem. Well, actually my 2.6.29 copy seems to be broken: arch/x86/kernel/amd_iommu.o: In function `debugfs_create_size_t': amd_iommu.c:(.text+0x0): multiple definition of `debugfs_create_size_t' arch/x86/kernel/kdebugfs.o:kdebugfs.c:(.text+0x0): first defined here make[1]: *** [arch/x86/kernel/built-in.o] Error 1 make: *** [arch/x86/kernel] Error 2 # git pull Already up-to-date. # git status # On branch master # Changed but not updated: # (use "git add <file>..." to update what will be committed) # (use "git checkout -- <file>..." to discard changes in working directory) # # modified: drivers/staging/winbond/wbusb.c # no changes added to commit (use "git add" and/or "git commit -a") I don't know too much how to use git to do what you're asking. I've some years of experience with svn and cvs but git is quite a new tool for me. I'll wait a git kernel that compile while I'll try to find out how to do what you're asking. I have not so much free time to spend in a checkout / build / test if the test phase need a kernel reinstall / system reboot. If it's only the driver module I think I can try to test it for each changeset in a couple of weeks /starting from the day I know how to bisect :-) ). - -- Sandro -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iEYEARECAAYFAkl+GWIACgkQV871CybFezlRXACfQTBrMm2hL9b+o5oU1/6/xZ+f gaYAnjyyawOFPBPIe/scq5pYupNrRLa3 =d1SV -----END PGP SIGNATURE----- ^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH] w35und: fix usb_control_msg() error handling in wb35_probe() 2009-01-26 20:13 ` Sandro Bonazzola @ 2009-01-26 20:22 ` Pekka Enberg 0 siblings, 0 replies; 4+ messages in thread From: Pekka Enberg @ 2009-01-26 20:22 UTC (permalink / raw) To: Sandro Bonazzola Cc: Pavel Machek, Greg Kroah-Hartman, linux-kernel, Linux Netdev List Sandro Bonazzola wrote: > Well, I don't know if this can be somehow related to a bug in the driver, but on > my system, after I've unplugged the USB trasnsitter, ksoftirqd/0 begin to use > 89% of my CPU time (top). I suspect that there are some issue that should be > fixed somewhere. Heh, yes, there's plenty of things to fix. :-) >> Actually, it's probably me who broke the driver. I think Greg merged >> most of my patches after 2.6.28 came out. Sandro, if you have some >> time, please consider doing git bisect to find out which changeset >> causes the problem. > > Well, actually my 2.6.29 copy seems to be broken: > arch/x86/kernel/amd_iommu.o: In function `debugfs_create_size_t': > amd_iommu.c:(.text+0x0): multiple definition of `debugfs_create_size_t' > arch/x86/kernel/kdebugfs.o:kdebugfs.c:(.text+0x0): first defined here > make[1]: *** [arch/x86/kernel/built-in.o] Error 1 > make: *** [arch/x86/kernel] Error 2 > > # git pull > Already up-to-date. > > # git status > # On branch master > # Changed but not updated: > # (use "git add <file>..." to update what will be committed) > # (use "git checkout -- <file>..." to discard changes in working directory) > # > # modified: drivers/staging/winbond/wbusb.c > # > no changes added to commit (use "git add" and/or "git commit -a") > > > I don't know too much how to use git to do what you're asking. I've some years > of experience with svn and cvs but git is quite a new tool for me. > I'll wait a git kernel that compile while I'll try to find out how to do what > you're asking. Here's a tutorial on git bisect. http://kernel.org/pub/software/scm/git/docs/v1.3.3/howto/isolate-bugs-with-bisect.txt > I have not so much free time to spend in a checkout / build / test if the test > phase need a kernel reinstall / system reboot. If it's only the driver module I > think I can try to test it for each changeset in a couple of weeks /starting > from the day I know how to bisect :-) ). Assuming 2.6.29-rc1 is broken, you initialize the bisection with: git bisect start -- drivers/staging/winbond/ git bisect good v2.6.28 git bisect bad v2.6.29-rc1 Then you continue doing this: <apply my patch, compile, test> and git bisect good or git bisect bad depending on whether the kernel worked or not. Repeat until git bisect tells you which changeset broke the kernel. The process shouldn't take more than 5 compiles/reboots. Pekka ^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2009-01-26 20:26 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <Pine.LNX.4.64.0901221205400.27341@melkki.cs.Helsinki.FI>
[not found] ` <20090122114237.GA13845@atrey.karlin.mff.cuni.cz>
[not found] ` <4978C812.90702@gmail.com>
[not found] ` <20090122211052.GA1928@elf.ucw.cz>
[not found] ` <4978F47C.6010909@gmail.com>
[not found] ` <4978F3CF.2010301@cs.helsinki.fi>
[not found] ` <497E0F94.8000003@gmail.com>
2009-01-26 19:40 ` [PATCH] w35und: fix usb_control_msg() error handling in wb35_probe() Pekka Enberg
2009-01-26 19:43 ` Pekka Enberg
2009-01-26 20:13 ` Sandro Bonazzola
2009-01-26 20:22 ` Pekka Enberg
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox