Bug 1467952 - valgrind: Mask CPUID support in HWCAP on aarch64
valgrind: Mask CPUID support in HWCAP on aarch64
Status: ON_QA
Product: Red Hat Developer Toolset
Classification: Red Hat
Component: valgrind (Show other bugs)
DTS 7.0 RHEL 7
Unspecified Unspecified
unspecified Severity unspecified
: alpha
: 7.0
Assigned To: Mark Wielaard
:
Depends On: 1464085 1464211
Blocks:
  Show dependency treegraph
 
Reported: 2017-07-05 11:21 EDT by Mark Wielaard
Modified: 2017-08-03 11:56 EDT (History)
8 users (show)

See Also:
Fixed In Version: devtoolset-7-valgrind-3.13.0-4
Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: 1464085
Environment:
Last Closed:
Type: Bug
Regression: ---
Mount Type: ---
Documentation: ---
CRM:
Verified Versions:
Category: ---
oVirt Team: ---
RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: ---


Attachments (Terms of Use)


External Trackers
Tracker ID Priority Status Summary Last Updated
KDE Software Compilation 381556 None None None 2017-07-05 11:21 EDT

  None (edit)
Description Mark Wielaard 2017-07-05 11:21:27 EDT
+++ This bug was initially created as a clone of Bug #1464085 +++

valgrind currently does not know anything about the CPUID flag added to the HWCAP auxv entry in kernel 4.11.  It passes this flag through to applications, but it will then choke when the application uses it, like this:

ARM64 front end: branch_etc
disInstr(arm64): unhandled instruction 0xD5380000
disInstr(arm64): 1101'0101 0011'1000 0000'0000 0000'0000
==924== valgrind: Unrecognised instruction at address 0x11f548.
==924==    at 0x11F548: init_cpu_features (cpu-features.c:32)
==924==    by 0x11F548: dl_platform_init (dl-machine.h:241)
==924==    by 0x11F548: _dl_sysdep_start (dl-sysdep.c:231)
==924==    by 0x10981B: _dl_start_final (rtld.c:412)
==924==    by 0x109AAB: _dl_start (rtld.c:520)

This is from a newer glibc (not the one in Red Hat Enterprise Linux).  The crashing instruction is the mrs in the glibc startup code:

  if (hwcap & HWCAP_CPUID)
    {
      register uint64_t id = 0;
      asm volatile ("mrs %0, midr_el1" : "=r"(id));
      cpu_features->midr_el1 = id;
    }
  else
    cpu_features->midr_el1 = 0;

Perhaps valgrind should mask all the HWCAP bits it knows nothing about.

--- Additional comment from Florian Weimer on 2017-06-22 12:21:40 EDT ---

Workaround: Run with “LD_HWCAP_MASK=1”.

--- Additional comment from Mark Wielaard on 2017-06-23 06:55:45 EDT ---

The upstream bug is https://bugs.kde.org/show_bug.cgi?id=381556
arm64: Handle feature registers access on 4.11 Linux kernel or later

Fedora valgrind-3.13.0-3.fc27 contains a workaround, see bug https://bugzilla.redhat.com/show_bug.cgi?id=1464211
Comment 2 Mark Wielaard 2017-07-05 11:33:12 EDT
Note that this is nor urgent till there is a glibc update that relies on the auxv HWCAP setting on arm64. But programs on arm64 might check the HWCAP themselves and then use some instructions that valgrind doesn't support.

Easiest to see what the HWCAP is with/without valgrind is to run:

LD_SHOW_AUXV=1 valgrind -q /bin/true | grep HWCAP

Which will show the auxv twice, once for valgrind itself, then for /bin/true running under valgrind.

Note You need to log in before you can comment on or make changes to this bug.