[PATCH v2 0/2] sigaltstack: SS_AUTODISARM fixes

[PATCH v2 0/2] sigaltstack: SS_AUTODISARM fixes

[Stas Sergeev]

This patch set adds the missing SS_AUTODISARM handling for
compatibility mode (CONFIG_COMPAT). It is needed for dosemu
to not crash when built with -m32.

CC: deller@gmx.de
CC: linux-kernel@vger.kernel.org
CC: Andy Lutomirski <luto@kernel.org>

Stas Sergeev (2):
  sigaltstack: support SS_AUTODISARM for CONFIG_COMPAT
  tests: improve output of sigaltstack testcase

 include/linux/compat.h                    |  4 +++-
 kernel/signal.c                           | 11 +++++++++--
 tools/testing/selftests/sigaltstack/sas.c |  7 ++++---
 3 files changed, 16 insertions(+), 6 deletions(-)

-- 
2.9.3

Re: [PATCH 1/2] sigaltstack: support SS_AUTODISARM for CONFIG_COMPAT

[Andy Lutomirski]

On Sun, Feb 5, 2017 at 2:12 AM, Stas Sergeev <stsp@list.ru> wrote:
> Currently SS_AUTODISARM is not supported in compatibility
> mode, but does not return -EINVAL either. This makes dosemu
> built with -m32 on x86_64 to crash. Also the kernel's sigaltstack
> selftest fails if compiled with -m32.
>
> This patch adds the needed support.
>
> Suggesting for -stable.

This looks correct.  I find it odd that we used to check the old sp to
figure out whether to put SA_ONSTACK into the signal frame, but I do
think the new behavior is better.  In any event, this changes the
compat case to work like the native case, and no one noticed when the
native behavior changed, so it's fine by me.

--Andy

Re: [PATCH 1/2] sigaltstack: support SS_AUTODISARM for CONFIG_COMPAT

[Stas Sergeev]

05.02.2017 21:31, Andy Lutomirski пишет:
> On Sun, Feb 5, 2017 at 2:12 AM, Stas Sergeev <stsp@list.ru> wrote:
>> Currently SS_AUTODISARM is not supported in compatibility
>> mode, but does not return -EINVAL either. This makes dosemu
>> built with -m32 on x86_64 to crash. Also the kernel's sigaltstack
>> selftest fails if compiled with -m32.
>>
>> This patch adds the needed support.
>>
>> Suggesting for -stable.
> This looks correct.
So it seems my patches haven't made it into LKML
and I don't know if they made it into stable@ because
the archive link:
http://dir.gmane.org/gmane.linux.kernel.stable
doesn't work.
There is definitely something wrong here.

So did they pass into a stable ML?
If not, Andy can I use your Acked-by when/if resending?
Are there any known problems with MLs and the archives now?

Re: [PATCH 1/2] sigaltstack: support SS_AUTODISARM for CONFIG_COMPAT

[Andrew Morton]

On Mon, 6 Feb 2017 12:30:38 +0300 Stas Sergeev <stsp@list.ru> wrote:

> 05.02.2017 21:31, Andy Lutomirski __________:
> > On Sun, Feb 5, 2017 at 2:12 AM, Stas Sergeev <stsp@list.ru> wrote:
> >> Currently SS_AUTODISARM is not supported in compatibility
> >> mode, but does not return -EINVAL either. This makes dosemu
> >> built with -m32 on x86_64 to crash. Also the kernel's sigaltstack
> >> selftest fails if compiled with -m32.
> >>
> >> This patch adds the needed support.
> >>
> >> Suggesting for -stable.
> > This looks correct.
> So it seems my patches haven't made it into LKML
> and I don't know if they made it into stable@ because
> the archive link:
> http://dir.gmane.org/gmane.linux.kernel.stable
> doesn't work.
> There is definitely something wrong here.
> 
> So did they pass into a stable ML?
> If not, Andy can I use your Acked-by when/if resending?
> Are there any known problems with MLs and the archives now?

I have a copy of [1/2] because I was on cc.  I cannot locate a copy of
[2/2].  Please cc Oleg on signal-related changes.

[PATCH 2/2] tests: improve output of sigaltstack testcase

[Stas Sergeev]

[-- Attachment #1: Type: text/plain, Size: 1246 bytes --]

07.02.2017 03:04, Andrew Morton пишет:
> On Mon, 6 Feb 2017 12:30:38 +0300 Stas Sergeev <stsp@list.ru> wrote:
>
>> So it seems my patches haven't made it into LKML
>> and I don't know if they made it into stable@ because
>> the archive link:
>> http://dir.gmane.org/gmane.linux.kernel.stable
>> doesn't work.
>> There is definitely something wrong here.
>>
>> So did they pass into a stable ML?
>> If not, Andy can I use your Acked-by when/if resending?
>> Are there any known problems with MLs and the archives now?
> I have a copy of [1/2] because I was on cc.  I cannot locate a copy of
> [2/2].  Please cc Oleg on signal-related changes.
Thanks for taking these!
2/2 had the CCs to

CC: Shuah Khan<shuah@kernel.org>
CC: Ingo Molnar<mingo@kernel.org>
CC: Andy Lutomirski<luto@kernel.org>
CC:linux-kselftest@vger.kernel.org
CC:linux-kernel@vger.kernel.org
CC:trivial@kernel.org

as it only changed the self-test.
Obviously I am getting spam-blocked somewhere.
I generate the CC list with get_maintainers.pl. Is this
the right thing to do?
Oleg was CCed on [1/2] as get_maintainers.pl mentioned
him.

I am attaching the [2/2] to this e-mail.
If this doesn't work, I can try to re-send it with more
CCs, but it will likely be blocked again.

[-- Attachment #2: 0002-tests-improve-output-of-sigaltstack-testcase.patch --]
[-- Type: text/x-patch, Size: 1902 bytes --]

>From d4dc2ac235b5a18f5bdf916152509f8b0e11276e Mon Sep 17 00:00:00 2001
From: Stas Sergeev <stsp@list.ru>
Date: Sat, 4 Feb 2017 22:11:49 +0300
Subject: [PATCH 2/2] tests: improve output of sigaltstack testcase

Currently it uses %i for bitmasks, which makes it difficult
to properly decode the values. Use %x instead.

Signed-off-by: Stas Sergeev <stsp@users.sourceforge.net>
CC: Shuah Khan <shuah@kernel.org>
CC: Ingo Molnar <mingo@kernel.org>
CC: Andy Lutomirski <luto@kernel.org>
CC: linux-kselftest@vger.kernel.org
CC: linux-kernel@vger.kernel.org
CC: trivial@kernel.org
---
 tools/testing/selftests/sigaltstack/sas.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/tools/testing/selftests/sigaltstack/sas.c b/tools/testing/selftests/sigaltstack/sas.c
index 1bb0125..ccd0734 100644
--- a/tools/testing/selftests/sigaltstack/sas.c
+++ b/tools/testing/selftests/sigaltstack/sas.c
@@ -57,7 +57,7 @@ void my_usr1(int sig, siginfo_t *si, void *u)
 		exit(EXIT_FAILURE);
 	}
 	if (stk.ss_flags != SS_DISABLE)
-		printf("[FAIL]\tss_flags=%i, should be SS_DISABLE\n",
+		printf("[FAIL]\tss_flags=%x, should be SS_DISABLE\n",
 				stk.ss_flags);
 	else
 		printf("[OK]\tsigaltstack is disabled in sighandler\n");
@@ -122,7 +122,8 @@ int main(void)
 	if (stk.ss_flags == SS_DISABLE) {
 		printf("[OK]\tInitial sigaltstack state was SS_DISABLE\n");
 	} else {
-		printf("[FAIL]\tInitial sigaltstack state was %i; should have been SS_DISABLE\n", stk.ss_flags);
+		printf("[FAIL]\tInitial sigaltstack state was %x; "
+		       "should have been SS_DISABLE\n", stk.ss_flags);
 		return EXIT_FAILURE;
 	}
 
@@ -165,7 +166,7 @@ int main(void)
 		exit(EXIT_FAILURE);
 	}
 	if (stk.ss_flags != SS_AUTODISARM) {
-		printf("[FAIL]\tss_flags=%i, should be SS_AUTODISARM\n",
+		printf("[FAIL]\tss_flags=%x, should be SS_AUTODISARM\n",
 				stk.ss_flags);
 		exit(EXIT_FAILURE);
 	}
-- 
2.9.3