linux-kernel - Re: [PATCHv9 2/2] selftest/x86: add mremap vdso test

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: <20160617080346.GB30525@gmail.com>
Date:	Fri, 17 Jun 2016 10:03:46 +0200
From:	Ingo Molnar <mingo@...nel.org>
To:	Dmitry Safonov <dsafonov@...tuozzo.com>
Cc:	linux-kernel@...r.kernel.org, mingo@...hat.com,
	luto@...capital.net, tglx@...utronix.de, hpa@...or.com,
	x86@...nel.org, akpm@...ux-foundation.org, linux-mm@...ck.org,
	0x7f454c46@...il.com, Shuah Khan <shuahkh@....samsung.com>,
	linux-kselftest@...r.kernel.org
Subject: Re: [PATCHv9 2/2] selftest/x86: add mremap vdso test


* Dmitry Safonov <dsafonov@...tuozzo.com> wrote:

> Should print on success:
> [root@...alhost ~]# ./test_mremap_vdso_32
> 	AT_SYSINFO_EHDR is 0xf773f000
> [NOTE]	Moving vDSO: [f773f000, f7740000] -> [a000000, a001000]
> [OK]
> Or segfault if landing was bad (before patches):
> [root@...alhost ~]# ./test_mremap_vdso_32
> 	AT_SYSINFO_EHDR is 0xf774f000
> [NOTE]	Moving vDSO: [f774f000, f7750000] -> [a000000, a001000]
> Segmentation fault (core dumped)

Yeah, so I changed my mind again, I still don't like that the testcase faults on 
old kernels:

 triton:~/tip/tools/testing/selftests/x86> ./test_mremap_vdso_32 
         AT_SYSINFO_EHDR is 0xf7786000
 [NOTE]  Moving vDSO: [0xf7786000, 0xf7787000] -> [0xf7781000, 0xf7782000]
 Segmentation fault

How do I know that this testcase is special and that a segmentation fault in this 
case means that I'm running it on a too old kernel and that it's not some other 
unexpected failure in the test?

At minimum please run it behind fork() and catch the -SIGSEGV child exit:

  mremap(0xf7747000, 4096, 4096, MREMAP_MAYMOVE|MREMAP_FIXED, 0xf7742000) = 0xf7742000
  --- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0xf7747be9} ---
  +++ killed by SIGSEGV +++

and print:

  [FAIL] mremap() of the vDSO does not work on this kernel!

or such.

Ok?

Thanks,

	Ingo