On Wed, Mar 25, 2015 at 5:21 AM, Ingo Molnar <mi...@kernel.org> wrote: > > * Andy Lutomirski <l...@amacapital.net> wrote: > >> On Tue, Mar 24, 2015 at 1:08 PM, Ingo Molnar <mi...@kernel.org> wrote: >> > >> > * Andy Lutomirski <l...@kernel.org> wrote: >> > >> >> We currently have a race: if we're preempted during syscall exit, we >> >> can fail to process syscall return work that is queued up while >> >> we're preempted in ret_from_sys_call after checking ti.flags. >> >> >> >> Fix it by disabling interrupts before checking ti.flags. >> >> >> >> Fixes: 96b6352c1271 x86_64, entry: Remove the syscall exit audit and >> >> schedule optimizations >> >> Reported-by: Stefan Seyfried <stefan.seyfr...@googlemail.com> >> >> Reported-by: Takashi Iwai <ti...@suse.de> >> >> Signed-off-by: Andy Lutomirski <l...@kernel.org> >> >> --- >> >> >> >> Ingo, I don't understand the LOCKDEP_SYS_EXIT stuff. Can you take a quick >> >> look to confirm that it's okay to call it more than once? >> > >> > So the essence is that it wants to print this warning if we are >> > holding a lock after a syscall: >> > >> > printk("[ BUG: lock held when returning to user space! >> > ]\n"); >> > >> > it manipulates no state and is not sensitive to whether it's called >> > before or after return-work processing. >> > >> >> arch/x86/kernel/entry_64.S | 18 ++++++++++++++---- >> >> 1 file changed, 14 insertions(+), 4 deletions(-) >> >> >> >> diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S >> >> index 1d74d161687c..2babb393915e 100644 >> >> --- a/arch/x86/kernel/entry_64.S >> >> +++ b/arch/x86/kernel/entry_64.S >> >> @@ -364,12 +364,21 @@ system_call_fastpath: >> >> * Has incomplete stack frame and undefined top of stack. >> >> */ >> >> ret_from_sys_call: >> >> - testl $_TIF_ALLWORK_MASK,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) >> >> - jnz int_ret_from_sys_call_fixup /* Go the the slow path */ >> >> - >> >> LOCKDEP_SYS_EXIT >> >> DISABLE_INTERRUPTS(CLBR_NONE) >> >> TRACE_IRQS_OFF >> >> + >> >> + /* >> >> + * We must check ti flags with interrupts (or at least preemption) >> >> + * off because we must *never* return to userspace without >> >> + * processing exit work that is enqueued if we're preempted here. >> >> + * In particular, returning to userspace with any of the one-shot >> >> + * flags (TIF_NOTIFY_RESUME, TIF_USER_RETURN_NOTIFY, etc) set is >> >> + * very bad. >> >> + */ >> >> + testl $_TIF_ALLWORK_MASK,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) >> >> + jnz int_ret_from_sys_call_fixup /* Go the the slow path */ >> > >> > Should be safe to call it once again after user-work processing has >> > been finished. >> > >> > I've picked up your fix for tip:x86/urgent. >> >> FWIW, the tentative merge here: >> >> https://git.kernel.org/cgit/linux/kernel/git/tip/tip.git/commit/?h=tmp.tmp&id=a77dd1607ad88a601259a74ba4d646fa68b7cd9a >> >> looks funny. Why aren't you jumping to int_ret_from_sys_call_irqs_off? > > Indeed - the orphaned label should have told me that. The mismerge is > functionally harmless (causes extra overhead in the slowpath), that's > why it passed testing. > > Does: > > 06ab9c1ba6a1 Merge branch 'x86/urgent' into x86/asm, to resolve conflict > > look better to you?
Yes, looks good. Thanks. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/