Re: 2.6.21-mm2 -- compat_core_sys_select() kmalloc(0)

[Andy Whitcroft]

Seems that compat_core_sys_select() is using kmalloc(0) and SLUB is
picking up on it.  Occuring on an older x86_64 numa system.

BUG: at include/linux/slub_def.h:88 kmalloc_index()

Call Trace:
 [<ffffffff80270392>] get_slab+0x50/0x1f7
 [<ffffffff802706d0>] __kmalloc+0xf/0x6d
 [<ffffffff802a0421>] compat_core_sys_select+0x7e/0x1d8
 [<ffffffff802a061d>] compat_sys_select+0xa2/0x131
 [<ffffffff8021acb2>] ia32_sysret+0x0/0xa


This appears to occur when a 32 bit app uses select(0, ...) as a
timeout.  I think its safe as it is, in that we'll never touch the
pointers if n==0 but more eyes would help.  It might be clearer to
special case the n==0 case to avoid all this.

-apw
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/