linux-kernel - Re: [PATCH v4 2/2] selftests: Add a test mangling with uc

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <09370b2d-2b52-4133-8e0d-83e66b801ff8@arm.com>
Date: Mon, 15 Jul 2024 17:19:03 +0530
From: Dev Jain <dev.jain@....com>
To: Oleg Nesterov <oleg@...hat.com>
Cc: shuah@...nel.org, mingo@...nel.org, tglx@...utronix.de,
 mark.rutland@....com, ryan.roberts@....com, broonie@...nel.org,
 suzuki.poulose@....com, Anshuman.Khandual@....com,
 DeepakKumar.Mishra@....com, aneesh.kumar@...nel.org,
 linux-kselftest@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v4 2/2] selftests: Add a test mangling with uc_sigmask


On 6/30/24 20:48, Oleg Nesterov wrote:
> I see nothing wrong, but perhaps this test can be simplified?
> Feel free to ignore.
>
> Say,
>
> On 06/27, Dev Jain wrote:
>> +void handler_usr(int signo, siginfo_t *info, void *uc)
>> +{
>> +	int ret;
>> +
>> +	/*
>> +	 * Break out of infinite recursion caused by raise(SIGUSR1) invoked
>> +	 * from inside the handler
>> +	 */
>> +	++cnt;
>> +	if (cnt > 1)
>> +		return;
>> +
>> +	ksft_print_msg("In handler_usr\n");
>> +
>> +	/* SEGV blocked during handler execution, delivered on return */
>> +	if (raise(SIGSEGV))
>> +		ksft_exit_fail_perror("raise");
>> +
>> +	ksft_print_msg("SEGV bypassed successfully\n");
> You could simply do sigprocmask(SIG_SETMASK, NULL, &oldset) and check if
> SIGSEGV is blocked in oldset. SIG_SETMASK has no effect if newset == NULL.
>

IMHO, isn't raising the signal, and the process not terminating, a 
stricter test? I have already included your described approach in
the last testcase; so, the test includes both ways: raising the
signal -> process not terminating, and checking blockage with sigprocmask().