linux/arch
David Carrillo-Cisneros 19fc9ddd61 perf/x86/intel: Fix MSR_LAST_BRANCH_FROM_x bug when no TSX
Intel's SDM states that bits 61:62 in MSR_LAST_BRANCH_FROM_x are the
TSX flags for formats with LBR_TSX flags (i.e. LBR_FORMAT_EIP_EFLAGS2).

However, when the CPU has TSX support deactivated, bits 61:62 actually
behave as follows:

  - For wrmsr(), bits 61:62 are considered part of the sign extension.
  - When capturing branches, the LBR hw will always clear bits 61:62.
    regardless of the sign extension.

Therefore, if:

  1) LBR has TSX format.
  2) CPU has no TSX support enabled.

... then any value passed to wrmsr() must be sign extended to 63 bits
and any value from rdmsr() must be converted to have a sign extension
of 61 bits, ignoring the values at TSX flags.

This bug was masked by the work-around to the Intel's CPU bug:
BJ94. "LBR May Contain Incorrect Information When Using FREEZE_LBRS_ON_PMI"
in Document Number: 324643-037US.

The aforementioned work-around uses hw flags to filter out all kernel
branches, limiting LBR callstack to user level execution only.

Since user addresses are not sign extended, they do not trigger the wrmsr()
bug in MSR_LAST_BRANCH_FROM_x when saved/restored at context switch.

To verify the hw bug:

  $ perf record -b -e cycles sleep 1
  $ rdmsr -p 0 0x680
  0x1fffffffb0b9b0cc
  $ wrmsr -p 0 0x680 0x1fffffffb0b9b0cc
  write(): Input/output error

The quirk for LBR_FROM_ MSRs is required before calls to wrmsrl() and
after rdmsrl().

This patch introduces it for wrmsrl()'s done for testing LBR support.

Future patch in series adds the quirk for context switch, that would
be required if LBR callstack is to be enabled for ring 0.

Signed-off-by: David Carrillo-Cisneros <davidcc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Stephane Eranian <eranian@google.com>
Reviewed-by: Andi Kleen <ak@linux.intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1466533874-52003-3-git-send-email-davidcc@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2016-06-27 11:34:19 +02:00
..
alpha tree wide: get rid of __GFP_REPEAT for order-0 allocations part I 2016-06-24 17:23:52 -07:00
arc arc: get rid of superfluous __GFP_REPEAT 2016-06-24 17:23:52 -07:00
arm tree wide: get rid of __GFP_REPEAT for order-0 allocations part I 2016-06-24 17:23:52 -07:00
arm64 Merge branch 'akpm' (patches from Andrew) 2016-06-24 19:08:33 -07:00
avr32 tree wide: get rid of __GFP_REPEAT for order-0 allocations part I 2016-06-24 17:23:52 -07:00
blackfin printk/nmi: generic solution for safe printk in NMI 2016-05-20 17:58:30 -07:00
c6x asm-generic patch for 4.7 2016-05-24 15:24:37 -07:00
cris tree wide: get rid of __GFP_REPEAT for order-0 allocations part I 2016-06-24 17:23:52 -07:00
frv tree wide: get rid of __GFP_REPEAT for order-0 allocations part I 2016-06-24 17:23:52 -07:00
h8300 Merge branch 'hash' of git://ftp.sciencehorizons.net/linux 2016-05-28 16:15:25 -07:00
hexagon tree wide: get rid of __GFP_REPEAT for order-0 allocations part I 2016-06-24 17:23:52 -07:00
ia64 fix up initial thread stack pointer vs thread_info confusion 2016-06-24 17:07:33 -07:00
m32r Merge branch 'kbuild' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild 2016-05-26 22:01:22 -07:00
m68k tree wide: get rid of __GFP_REPEAT for order-0 allocations part I 2016-06-24 17:23:52 -07:00
metag tree wide: get rid of __GFP_REPEAT for order-0 allocations part I 2016-06-24 17:23:52 -07:00
microblaze tree wide: get rid of __GFP_REPEAT for order-0 allocations part I 2016-06-24 17:23:52 -07:00
mips mips: get rid of superfluous __GFP_REPEAT 2016-06-24 17:23:52 -07:00
mn10300 Merge branch 'akpm' (patches from Andrew) 2016-06-24 19:08:33 -07:00
nios2 nios2: get rid of superfluous __GFP_REPEAT 2016-06-24 17:23:52 -07:00
openrisc tree wide: get rid of __GFP_REPEAT for order-0 allocations part I 2016-06-24 17:23:52 -07:00
parisc parisc: get rid of superfluous __GFP_REPEAT 2016-06-24 17:23:52 -07:00
powerpc powerpc fixes for 4.7 #4 2016-06-25 06:01:48 -07:00
s390 s390: get rid of superfluous __GFP_REPEAT 2016-06-24 17:23:52 -07:00
score score: get rid of superfluous __GFP_REPEAT 2016-06-24 17:23:52 -07:00
sh sh: get rid of superfluous __GFP_REPEAT 2016-06-24 17:23:52 -07:00
sparc sparc: get rid of superfluous __GFP_REPEAT 2016-06-24 17:23:52 -07:00
tile Merge branch 'akpm' (patches from Andrew) 2016-06-24 19:08:33 -07:00
um tree wide: get rid of __GFP_REPEAT for order-0 allocations part I 2016-06-24 17:23:52 -07:00
unicore32 unicore32: get rid of superfluous __GFP_REPEAT 2016-06-24 17:23:52 -07:00
x86 perf/x86/intel: Fix MSR_LAST_BRANCH_FROM_x bug when no TSX 2016-06-27 11:34:19 +02:00
xtensa tree wide: get rid of __GFP_REPEAT for order-0 allocations part I 2016-06-24 17:23:52 -07:00
.gitignore
Kconfig Clarify naming of thread info/stack allocators 2016-06-24 15:09:37 -07:00