[PATCH] x86/entry/32: Fix entry_INT80_32 to expect interrupts to be on

[PATCH] x86/entry/32: Fix entry_INT80_32 to expect interrupts to be on

[Andy Lutomirski]

When I rewrote entry_INT80_32, I thought that int80 was an interrupt
gate.  It's a trap gate.  *facepalm*

Thanks to Brian Gerst for pointing out that it's better to change
the entry code than to change the gate type.

Suggested-by: Brian Gerst <brgerst@gmail.com>
Reported-by: Borislav Petkov <bp@suse.de>
Fixes: 150ac78d63af ("x86/entry/32: Switch INT80 to the new C syscall path")
Signed-off-by: Andy Lutomirski <luto@kernel.org>
---

I can reproduce Borislav's warning with an appropriate .config, and this
patch fixes it.  I removed the Tested-by, because it's not the same fix
that Borislav tested.

 arch/x86/entry/common.c   | 9 +--------
 arch/x86/entry/entry_32.S | 8 ++++----
 2 files changed, 5 insertions(+), 12 deletions(-)

diff --git a/arch/x86/entry/common.c b/arch/x86/entry/common.c
index b53e04d301a3..b0293e0a60df 100644
--- a/arch/x86/entry/common.c
+++ b/arch/x86/entry/common.c
@@ -351,7 +351,7 @@ __visible inline void syscall_return_slowpath(struct pt_regs *regs)
  * in workloads that use it, and it's usually called from
  * do_fast_syscall_32, so forcibly inline it to improve performance.
  */
-static __always_inline void do_syscall_32_irqs_on(struct pt_regs *regs)
+__visible __always_inline void do_syscall_32_irqs_on(struct pt_regs *regs)
 {
 	struct thread_info *ti = pt_regs_to_thread_info(regs);
 	unsigned int nr = (unsigned int)regs->orig_ax;
@@ -386,13 +386,6 @@ static __always_inline void do_syscall_32_irqs_on(struct pt_regs *regs)
 	syscall_return_slowpath(regs);
 }
 
-/* Handles int $0x80 */
-__visible void do_int80_syscall_32(struct pt_regs *regs)
-{
-	local_irq_enable();
-	do_syscall_32_irqs_on(regs);
-}
-
 /* Returns 0 to return using IRET or 1 to return using SYSEXIT/SYSRETL. */
 __visible long do_fast_syscall_32(struct pt_regs *regs)
 {
diff --git a/arch/x86/entry/entry_32.S b/arch/x86/entry/entry_32.S
index 07a041b91af1..ba0dad937be9 100644
--- a/arch/x86/entry/entry_32.S
+++ b/arch/x86/entry/entry_32.S
@@ -346,13 +346,13 @@ ENTRY(entry_INT80_32)
 	SAVE_ALL pt_regs_ax=$-ENOSYS	/* save rest */
 
 	/*
-	 * User mode is traced as though IRQs are on, and the interrupt gate
-	 * turned them off.
+	 * User mode is traced as though IRQs are on.  Unlike the 64-bit
+	 * case, int80 is a trap gate on 32-bit kernels, so interrupts
+	 * are already on (unless user code is messing around with iopl).
 	 */
-	TRACE_IRQS_OFF
 
 	movl	%esp, %eax
-	call	do_int80_syscall_32
+	call	do_syscall_32_irqs_on
 .Lsyscall_32_done:
 
 restore_all:
-- 
2.4.3

Re: [PATCH] x86/entry/32: Fix entry_INT80_32 to expect interrupts to be on

[Borislav Petkov]

On Fri, Oct 16, 2015 at 11:47:11AM -0700, Andy Lutomirski wrote:
> When I rewrote entry_INT80_32, I thought that int80 was an interrupt
> gate.  It's a trap gate.  *facepalm*
> 
> Thanks to Brian Gerst for pointing out that it's better to change
> the entry code than to change the gate type.
> 
> Suggested-by: Brian Gerst <brgerst@gmail.com>
> Reported-by: Borislav Petkov <bp@suse.de>
> Fixes: 150ac78d63af ("x86/entry/32: Switch INT80 to the new C syscall path")
> Signed-off-by: Andy Lutomirski <luto@kernel.org>
> ---
> 
> I can reproduce Borislav's warning with an appropriate .config, and this
> patch fixes it.  I removed the Tested-by, because it's not the same fix
> that Borislav tested.

You can re-add it. Looks good.

-- 
Regards/Gruss,
    Boris.

ECO tip #101: Trim your mails when you reply.

Re: [PATCH] x86/entry/32: Fix entry_INT80_32 to expect interrupts to be on

[Andy Lutomirski]

On Fri, Oct 16, 2015 at 12:45 PM, Borislav Petkov <bp@alien8.de> wrote:
> On Fri, Oct 16, 2015 at 11:47:11AM -0700, Andy Lutomirski wrote:
>> When I rewrote entry_INT80_32, I thought that int80 was an interrupt
>> gate.  It's a trap gate.  *facepalm*

Ingo, Thomas, don't apply this.  It breaks the 64-bit build for stupid
reasons.  I'll send a v2.