linux-2.6.22-rc2. SLUB report[kzalloc(0)]
Hi.
There is a BUG message from SLUB during boot process:
--
May 21 05:39:10 midgard kernel: [ 31.177484] BUG: at include/linux/slub_def.h:77 kmalloc_index() May 21 05:39:10 midgard kernel: [ 31.178355] [<c01062ca>] show_trace_log_lvl+0x1a/0x30 May 21 05:39:10 midgard kernel: [ 31.179263] [<c0106e72>] show_trace +0x12/0x20 May 21 05:39:10 midgard kernel: [ 31.180177] [<c0106ee6>] dump_stack +0x16/0x20 May 21 05:39:10 midgard kernel: [ 31.181094] [<c018440d>] get_slab +0x1cd/0x260 May 21 05:39:10 midgard kernel: [ 31.182024] [<c0185809>] __kmalloc_track_caller+0x19/0xa0 May 21 05:39:10 midgard kernel: [ 31.183019] [<c0170059>] __kzalloc +0x19/0x50 May 21 05:39:10 midgard kernel: [ 31.184012] [<df05586e>] usb_get_configuration+0x9de/0x11c0 [usbcore] May 21 05:39:10 midgard kernel: [ 31.185115] [<df04cd07>] usb_new_device+0x17/0x190 [usbcore] May 21 05:39:10 midgard kernel: [ 31.186181] [<df04e68a>] hub_thread +0x79a/0xfd0 [usbcore] May 21 05:39:10 midgard kernel: [ 31.187185] [<c013b8e2>] kthread +0x42/0x70 May 21 05:39:10 midgard kernel: [ 31.188190] [<c0105ea7>]
kernel_thread_helper+0x7/0x10
--
kzalloc(0, GFP_KERNEL) occurs in drivers/usb/core/config.c in usb_get_configuration function. I already wrote about this slub bug report and offered a patch(http://www.uwsg.iu.edu/hypermail/linux/kernel/0705.1/0154.html). I don't know, may be it is not major thing, if it is, just ignore it. I just think, that
--
length = ncfg * sizeof(struct usb_host_config); dev->config = kzalloc(length, GFP_KERNEL);
--
isn't clear, because ncfg - in my case - is 0 and I suppose, that size of the leastest slab cache can become(in future) smaller than sizeof(struct usb_host_config).
Thanks for attention.
Signature · 1 line
Dan Kruchinin <[email protected]>
Dan Kruchinin napsal(a):
Quoted 37 lines
Hi.
There is a BUG message from SLUB during boot process:
--May 21 05:39:10 midgard kernel: [ 31.177484] BUG: at include/linux/slub_def.h:77 kmalloc_index() May 21 05:39:10 midgard kernel: [ 31.178355] [<c01062ca>] show_trace_log_lvl+0x1a/0x30 May 21 05:39:10 midgard kernel: [ 31.179263] [<c0106e72>] show_trace +0x12/0x20 May 21 05:39:10 midgard kernel: [ 31.180177] [<c0106ee6>] dump_stack +0x16/0x20 May 21 05:39:10 midgard kernel: [ 31.181094] [<c018440d>] get_slab +0x1cd/0x260 May 21 05:39:10 midgard kernel: [ 31.182024] [<c0185809>] __kmalloc_track_caller+0x19/0xa0 May 21 05:39:10 midgard kernel: [ 31.183019] [<c0170059>] __kzalloc +0x19/0x50 May 21 05:39:10 midgard kernel: [ 31.184012] [<df05586e>] usb_get_configuration+0x9de/0x11c0 [usbcore] May 21 05:39:10 midgard kernel: [ 31.185115] [<df04cd07>] usb_new_device+0x17/0x190 [usbcore] May 21 05:39:10 midgard kernel: [ 31.186181] [<df04e68a>] hub_thread +0x79a/0xfd0 [usbcore] May 21 05:39:10 midgard kernel: [ 31.187185] [<c013b8e2>] kthread +0x42/0x70 May 21 05:39:10 midgard kernel: [ 31.188190] [<c0105ea7>]
kernel_thread_helper+0x7/0x10
--
kzalloc(0, GFP_KERNEL)occurs indrivers/usb/core/config.cinusb_get_configurationfunction. I already wrote about this slub bug report and offered a patch(http://www.uwsg.iu.edu/hypermail/linux/kernel/0705.1/0154.html). I don't know, may be it is not major thing, if it is, just ignore it. I just think, that
--
Quoted 6 lines
length = ncfg * sizeof(struct usb_host_config); dev->config = kzalloc(length, GFP_KERNEL);--
isn't clear, because ncfg - in my case - is 0 and I suppose, that size of the leastest slab cache can become(in future) smaller thansizeof(struct usb_host_config).
There is yet another patch for this:
http://lkml.org/lkml/2007/5/19/171
sitting here:
http://kernel.org/pub/linux/kernel/people/gregkh/gregkh-2.6/gregkh-03-usb/usb-don-t-try-to-kzalloc-0-bytes.patch
regards,
Signature · 4 lines
http://www.fi.muni.cz/~xslaby/ Jiri Slaby faculty of informatics, masaryk university, brno, cz e-mail: jirislaby gmail com, gpg pubkey fingerprint: B674 9967 0407 CE62 ACC8 22A0 32CC 55C3 39D4 7A7E