Re: [PATCH 2/2] module: fix bne2 "gave up waiting for init ofmodule libcrc32c"

[Brandon Philips]

On 10:53 Tue 01 Jun 2010, Linus Torvalds wrote:
> On Tue, 1 Jun 2010, Linus Torvalds wrote:
> > So I do agree that your later two-patch series is the way to go.
> 
> Oh, btw, did we ever get that version tested by the only person who seems 
> to see the problem? Brandon?

Sorry for the delay. I broke the fuse on the remote power switch I was
using.

This is the subset of patches that everyone agreed, right?
 git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux-2.6 modules
Otherwise what patches should I be applying to 2.6.35-rc1?

When I tested a Kernel with Rusty's modules branch pulled onto
2.6.35-rc1 I got duplicate sysfs path errors:

[    8.985303] WARNING: at fs/sysfs/dir.c:451 sysfs_add_one+0xb2/0xd0()
[    8.998161] Hardware name: Dinar
[    9.004734] sysfs: cannot create duplicate filename '/module/libcrc32c'
[    9.018068] Modules linked in: libcrc32c(+) button ohci_hcd
ehci_hcd usbcore sd_mod crc_t10dif edd ext3 mbcache jbd fan thermal
processor thermal_sys hwmon ide_pci_generic atiixp ide_core
ata_generic ahci libahci pata_atiixp libata scsi_mod
[    9.062682] Pid: 1588, comm: modprobe Not tainted 2.6.35-rc1-0.7-default #3
[    9.076711] Call Trace:
[    9.081731]  [<ffffffff8116a162>] ? sysfs_add_one+0xb2/0xd0
[    9.092988]  [<ffffffff8104f6aa>] warn_slowpath_common+0x7a/0xd0
[    9.105115]  [<ffffffff8104f7a1>] warn_slowpath_fmt+0x41/0x50
[    9.116718]  [<ffffffff8116a162>] sysfs_add_one+0xb2/0xd0
[    9.127628]  [<ffffffff8116ace5>] create_dir+0x75/0xc0
[    9.138015]  [<ffffffff8116adc6>] sysfs_create_dir+0x96/0xc0
[    9.149447]  [<ffffffff811c5297>] kobject_add_internal+0xe7/0x250
[    9.161744]  [<ffffffff811c5518>] kobject_add_varg+0x38/0x60
[    9.173173]  [<ffffffff811c5593>] kobject_init_and_add+0x53/0x70
[    9.185295]  [<ffffffff811c5160>] ? kset_find_obj+0x40/0x90
[    9.196555]  [<ffffffff8107f193>] mod_sysfs_init+0x83/0xf0
[    9.207639]  [<ffffffff810807ec>] load_module+0xbec/0x1e10
[    9.218723]  [<ffffffff810d7467>] ? handle_mm_fault+0x227/0xaf0
[    9.230672]  [<ffffffff810da47c>] ? vma_link+0xac/0x110
[    9.241236]  [<ffffffff81385168>] ? do_page_fault+0x228/0x410
[    9.252839]  [<ffffffff810dcc19>] ? do_mmap_pgoff+0x359/0x380
[    9.264442]  [<ffffffff81081a6a>] sys_init_module+0x5a/0x220
[    9.275873]  [<ffffffff81002ec2>] system_call_fastpath+0x16/0x1b
[    9.287993] ---[ end trace 2a73785cc061234f ]---
[    9.297350] kobject_add_internal failed for libcrc32c with -EEXIST,
don't try to register things with the same name in the same directory.

To fix this we need to make sure that we only have one copy of a module
going through load_module at a time.  Patch suggestion follows which
boots without errors. I am sure there is a better way to do what it does
   ;)

Full dmesg:
  http://ifup.org/~philips/rusty-module-branch.dmesg

Cheers,

	Brandon