Re: [PATCH RESEND v3 6/6] powerpc/signal: Use...

Re: [PATCH RESEND v3 6/6] powerpc/signal: Use unsafe_copy_siginfo_to_user()

From: Eric W. Biederman <hidden>
Date: 2021-09-13 16:21:57

ebiederm@xmission.com (Eric W. Biederman) writes:

Christophe Leroy [off-list ref] writes:

quoted

Use unsafe_copy_siginfo_to_user() in order to do the copy
within the user access block.

On an mpc 8321 (book3s/32) the improvment is about 5% on a process
sending a signal to itself.

If you can't make function calls from an unsafe macro there is another
way to handle this that doesn't require everything to be inline.

From a safety perspective it is probably even a better approach.

Eric

diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c
index 0608581967f0..1b30bb78b863 100644
--- a/arch/powerpc/kernel/signal_32.c
+++ b/arch/powerpc/kernel/signal_32.c

@@ -205,6 +205,12 @@ struct sigframe {
 	int			abigap[56];
 };
 
+#ifdef CONFIG_PPC64
+# define user_siginfo_t		compat_siginfo_t
+#else
+# define user_siginfo_t		siginfo_t
+#endif
+
 /*
  *  When we have rt signals to deliver, we set up on the
  *  user stack, going down from the original stack pointer:

@@ -217,11 +223,7 @@ struct sigframe {
  *
  */
 struct rt_sigframe {
-#ifdef CONFIG_PPC64
-	compat_siginfo_t info;
-#else
-	struct siginfo info;
-#endif
+	user_siginfo_t info;
 	struct ucontext	uc;
 #ifdef CONFIG_PPC_TRANSACTIONAL_MEM
 	struct ucontext	uc_transact;

@@ -712,7 +714,7 @@ static long restore_tm_user_regs(struct pt_regs *regs, struct mcontext __user *s
 
 #ifdef CONFIG_PPC64
 
-#define copy_siginfo_to_user	copy_siginfo_to_user32
+#define copy_siginfo_to_external	copy_siginfo_to_external32
 
 #endif /* CONFIG_PPC64 */

@@ -731,6 +733,7 @@ int handle_rt_signal32(struct ksignal *ksig, sigset_t *oldset,
 	struct pt_regs *regs = tsk->thread.regs;
 	/* Save the thread's msr before get_tm_stackpointer() changes it */
 	unsigned long msr = regs->msr;
+	user_siginfo_t uinfo;
 
 	/* Set up Signal Frame */
 	frame = get_sigframe(ksig, tsk, sizeof(*frame), 1);

@@ -743,6 +746,8 @@ int handle_rt_signal32(struct ksignal *ksig, sigset_t *oldset,
 	else
 		prepare_save_user_regs(1);
 
+	copy_siginfo_to_external(&uinfo, &ksig->info);
+
 	if (!user_access_begin(frame, sizeof(*frame)))
 		goto badframe;

@@ -778,12 +783,10 @@ int handle_rt_signal32(struct ksignal *ksig, sigset_t *oldset,
 		asm("dcbst %y0; sync; icbi %y0; sync" :: "Z" (mctx->mc_pad[0]));
 	}
 	unsafe_put_sigset_t(&frame->uc.uc_sigmask, oldset, failed);
+	unsafe_copy_to_user(&frame->info, &uinfo, failed);
 
 	user_access_end();
 
-	if (copy_siginfo_to_user(&frame->info, &ksig->info))
-		goto badframe;
-
 	regs->link = tramp;
 
 #ifdef CONFIG_PPC_FPU_REGS

Eric

Re: [PATCH RESEND v3 6/6] powerpc/signal: Use unsafe_copy_siginfo_to_user()

From: Christophe Leroy <hidden>
Date: 2021-09-13 17:19:32


Le 13/09/2021 à 18:21, Eric W. Biederman a écrit :

ebiederm@xmission.com (Eric W. Biederman) writes:

quoted

Christophe Leroy [off-list ref] writes:

quoted

Use unsafe_copy_siginfo_to_user() in order to do the copy
within the user access block.

On an mpc 8321 (book3s/32) the improvment is about 5% on a process
sending a signal to itself.

If you can't make function calls from an unsafe macro there is another
way to handle this that doesn't require everything to be inline.

 From a safety perspective it is probably even a better approach.

Yes but that's exactly what I wanted to avoid for the native ppc32 case: 
this double hop means useless pressure on the cache. The siginfo_t 
structure is 128 bytes large, that means 8 lines of cache on powerpc 8xx.

But maybe it is acceptable to do that only for the compat case. Let me 
think about it, it might be quite easy.

Christophe

`h`	back out one level
`j`	next message in thread
`k`	previous message in thread
`l`	drill in
`Esc`	close help / fold thread tree
`?`	toggle this help