Re: [patch v6 12/12] btrace, x86: disable on some processors

["H.J. Lu" <hjl dot tools at gmail dot com>]

On Mon, Dec 17, 2012 at 8:01 AM,  <markus.t.metzger@intel.com> wrote:
> From: Markus Metzger <markus.t.metzger@intel.com>
>
> LBR, BTM, or BTS records may have incorrect branch "from" information afer an
> EIST transition, T-states, C1E, or Adaptive Thermal Throttling (AAJ122).
>
> This results in sporadic test fails. Disable btrace on those processors.
>
> 2012-12-17 Markus Metzger <markus.t.metzger@intel.com>
>
>         * common/linux-btrace.c (linux_supports_btrace): Add cpuid check.
>
>
> ---
>  gdb/common/linux-btrace.c |   43 +++++++++++++++++++++++++++++++++++++++++++
>  1 files changed, 43 insertions(+), 0 deletions(-)
>
> diff --git a/gdb/common/linux-btrace.c b/gdb/common/linux-btrace.c
> index 1231ccc..b5a41a5 100644
> --- a/gdb/common/linux-btrace.c
> +++ b/gdb/common/linux-btrace.c
> @@ -247,7 +247,50 @@ perf_event_read_bts (struct btrace_target_info* tinfo, const uint8_t *begin,
>  int
>  linux_supports_btrace (void)
>  {
> +#if defined(__i386__) || defined(__x86_64__)
> +  {
> +    unsigned int cpuid, model, family;
> +
> +    __asm__ __volatile__ ("movl   $1, %%eax;"
> +                         "cpuid;"
> +                         : "=a" (cpuid)
> +                         :: "%ebx", "%ecx", "%edx");
> +
> +    family = (cpuid >> 8) & 0xf;
> +    model = (cpuid >> 4) & 0xf;
>
> +    switch (family)
> +      {
> +      case 6:
> +       model += (cpuid >> 12) & 0xf0;

Do you need to verify that it is an Intel processor before
checking extended family/model?

> +       switch (model)
> +         {
> +         case 26: /* Nehalem */
> +         case 30:
> +         case 46:

I believe model 31 is also Nehalem.  BTW, most Intel documents
use hex number, not decimal for model numbers.

> +         case 37: /* Westmere */
> +         case 44:
> +         case 47:
> +         case 42: /* Sandy Bridge */
> +         case 45:
> +         case 58: /* Ivy Bridge */
> +
> +           /* AAJ122: LBR, BTM, or BTS records may have incorrect branch
> +              "from" information afer an EIST transition, T-states, C1E, or
> +              Adaptive Thermal Throttling.  */
> +           return 0;
> +         }
> +      }
> +  }
> +
>    return 1;
> +
> +#else /* defined(__i386__) || defined(__x86_64__) */
> +
> +  return 0;
> +
> +#endif /* defined(__i386__) || defined(__x86_64__) */
>  }
>
>  /* See linux-btrace.h.  */
> --
> 1.7.6.5
>



-- 
H.J.