Bug 16090
| Summary: | sysfs: cannot create duplicate filename | ||
|---|---|---|---|
| Product: | Drivers | Reporter: | Tobias (devnull) |
| Component: | PCI | Assignee: | drivers_pci (drivers_pci) |
| Status: | CLOSED CODE_FIX | ||
| Severity: | normal | CC: | brian.bloniarz, devnull, greg, jbarnes, m.s.tsirkin, maciej.rutecki, rjw |
| Priority: | P1 | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Kernel Version: | 2.6.35-rc1 | Subsystem: | |
| Regression: | Yes | Bisected commit-id: | |
| Bug Depends on: | |||
| Bug Blocks: | 16055 | ||
| Attachments: |
lspci -vv output on 2.6.34
.config for 3.6.35-rc2 (broken) |
||
|
Description
Tobias
2010-06-01 15:59:49 UTC
This is a pci core bug, not a sysfs one. I see this in 2.6.35-rc2 as well. 2.6.34 is fine. Created attachment 26680 [details]
lspci -vv output on 2.6.34
Created attachment 26681 [details]
.config for 3.6.35-rc2 (broken)
backtrace for me looks slightly different: ------------[ cut here ]------------ WARNING: at fs/sysfs/dir.c:451 sysfs_add_one+0xc9/0x130() Hardware name: X8DTN sysfs: cannot create duplicate filename '/devices/pci0000:00/0000:00:1c.4/0000:0a:00.0/slot' Modules linked in: Pid: 1, comm: swapper Not tainted 2.6.35-rc2-mst-debug #8 Call Trace: [<ffffffff8105ed2f>] warn_slowpath_common+0x7f/0xc0 [<ffffffff8105ee26>] warn_slowpath_fmt+0x46/0x50 [<ffffffff811ad249>] sysfs_add_one+0xc9/0x130 [<ffffffff811ad9d8>] sysfs_do_create_link+0xd8/0x1a0 [<ffffffff811adad3>] sysfs_create_link+0x13/0x20 [<ffffffff812353a6>] pci_create_sysfs_dev_files+0x3d6/0x5f0 [<ffffffff81c74b0f>] ? pci_sysfs_init+0x0/0x51 [<ffffffff81c74b0f>] ? pci_sysfs_init+0x0/0x51 [<ffffffff81c74b2e>] pci_sysfs_init+0x1f/0x51 [<ffffffff81c748e0>] ? pci_resource_alignment_sysfs_init+0x0/0x19 [<ffffffff8100204c>] do_one_initcall+0x3c/0x1d0 [<ffffffff81c458c1>] kernel_init+0x2b6/0x346 [<ffffffff8100be84>] kernel_thread_helper+0x4/0x10 [<ffffffff81c4560b>] ? kernel_init+0x0/0x346 [<ffffffff8100be80>] ? kernel_thread_helper+0x0/0x10 ---[ end trace 50015d47e04d483e ]--- Magic number: 14:612:577 Note: unsurprisingly, since the failing device is the disk, it fails to boot: 0a:00.0 IDE interface: JMicron Technology Corp. JMB368 IDE controller items from grub menu.lst: nothing special in kernel parameters:
title Red Hat Enterprise Linux (2.6.34-mst-debug)
root (hd0,0)
kernel /vmlinuz-2.6.34-mst-debug ro root=/dev/mapper/vg_virtlab16-lv_root rd_LVM_LV=vg_virtlab16/lv_root rd_NO_LUKS rd_NO_MD rd_NO_DM LANG=en_US.UTF-8 SYSFONT=latarcyrheb-sun16 KEYBOARDTYPE=pc KEYTABLE=us console=tty console=ttyS0,115200,n8 crashkernel=auto
initrd /initramfs-2.6.34-mst-debug.img
title Red Hat Enterprise Linux (2.6.35-rc2-mst-debug)
root (hd0,0)
kernel /vmlinuz-2.6.35-rc2-mst-debug ro root=/dev/mapper/vg_virtlab16-lv_root rd_LVM_LV=vg_virtlab16/lv_root rd_NO_LUKS rd_NO_MD rd_NO_DM LANG=en_US.UTF-8 SYSFONT=latarcyrheb-sun16 KEYBOARDTYPE=pc KEYTABLE=us console=tty console=ttyS0,115200,n8 crashkernel=auto
also reported here: http://lkml.org/lkml/2010/6/6/196 does reverting this commit resolve the problem? commit 75568f8094eb0333e9c2109b23cbc8b82d318a3c Author: Alex Chiang <achiang@hp.com> Date: Mon Mar 8 10:24:29 2010 -0700 PCI: create function symlinks in /sys/bus/pci/slots/N/ I did a git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6 and git revert 75568f8094eb0333e9c2109b23cbc8b82d318a3c, followed by my usual build routine. The resulting kernel doesn't show that error on boot anymore. Handled-By : Jesse Barnes <jbarnes@virtuousgeek.org> Please test 2.6.35-rc3, the patch you mention should be reverted: Author: Jesse Barnes <jbarnes@virtuousgeek.org> Date: Fri Jun 11 13:08:37 2010 -0700 Revert "PCI: create function symlinks in /sys/bus/pci/slots/N/" This reverts commit 75568f8094eb0333e9c2109b23cbc8b82d318a3c. Since they're just a convenience anyway, remove these symlinks since they're causing duplicate filename errors in the wild. Acked-by: Alex Chiang <achiang@canonical.com> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> Sorry, the commit id: commit 3be434f0244ee059432f92de7e891ee514f41738 -rc3 is fine. Thanks for confirming. |