From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Morton Date: Wed, 23 May 2007 06:43:14 +0000 Subject: Re: BUG: sleeping function called from invalid context at Message-Id: <20070522234314.54081817.akpm@linux-foundation.org> List-Id: References: <20070523063154.GR85884050@sgi.com> In-Reply-To: <20070523063154.GR85884050@sgi.com> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable To: David Chinner Cc: lkml , linux-ia64@vger.kernel.org On Wed, 23 May 2007 16:31:54 +1000 David Chinner wrote: > Saw this when running strace -f on a script on 2.6.21 on ia64: >=20 > BUG: sleeping function called from invalid context at kernel/fork.c:385 > in_atomic():1, irqs_disabled():0 >=20 > Call Trace: > [] show_stack+0x80/0xa0 > sp=E00000306e6f7a00 bsp=E00000306e6f0ef8 > [] dump_stack+0x30/0x60 > sp=E00000306e6f7bd0 bsp=E00000306e6f0ee0 > [] __might_sleep+0x1f0/0x260 > sp=E00000306e6f7bd0 bsp=E00000306e6f0eb8 > [] mmput+0x20/0x220 > sp=E00000306e6f7bd0 bsp=E00000306e6f0e90 > [] sys_ptrace+0x460/0x1600 > sp=E00000306e6f7bd0 bsp=E00000306e6f0d80 > [] ia64_ret_from_syscall+0x0/0x20 > sp=E00000306e6f7e30 bsp=E00000306e6f0d80 > [] __kernel_syscall_via_break+0x0/0x20 > sp=E00000306e6f8000 bsp=E00000306e6f0d80 > BUG: sleeping function called from invalid context at kernel/fork.c:385 > in_atomic():1, irqs_disabled():0 >=20 > Call Trace: > [] show_stack+0x80/0xa0 > sp=E00000306e6f7a00 bsp=E00000306e6f0ef8 > [] dump_stack+0x30/0x60 > sp=E00000306e6f7bd0 bsp=E00000306e6f0ee0 > [] __might_sleep+0x1f0/0x260 > sp=E00000306e6f7bd0 bsp=E00000306e6f0eb8 > [] mmput+0x20/0x220 > sp=E00000306e6f7bd0 bsp=E00000306e6f0e90 > [] sys_ptrace+0x460/0x1600 > sp=E00000306e6f7bd0 bsp=E00000306e6f0d80 > [] ia64_ret_from_syscall+0x0/0x20 > sp=E00000306e6f7e30 bsp=E00000306e6f0d80 > [] __kernel_syscall_via_break+0x0/0x20 > sp=E00000306e6f8000 bsp=E00000306e6f0d80 >=20 > I could reproduce it via 'strace -f sleep 1' >=20 I'd say this is specific to ia64. Someone would have spotted it on x86 by now.