Re: [PATCH 2/3] skip_prolgoue (amd64)

[Pedro Alves <palves@redhat.com>]

On 11/30/2013 02:33 AM, Yao Qi wrote:
> On 11/30/2013 12:05 AM, Mark Kettenis wrote:
>> But only if the cache is properly invalidated when control when
>> running/stepping.  Is it?
> 
> Yes, cache is invalidated when the inferior is to be resumed.  See
> target.c:target_resume.
> 
> void
> target_resume (ptid_t ptid, int step, enum gdb_signal signal)
> {
>    struct target_ops *t;
> 
>    target_dcache_invalidate ();
>    ....
> 

We also invalidate the cache just before running a command, threads
may be running, in prepare_execute_command:

struct cleanup *
prepare_execute_command (void)
{
...
  /* With multiple threads running while the one we're examining is
     stopped, the dcache can get stale without us being able to detect
     it.  For the duration of the command, though, use the dcache to
     help things like backtrace.  */
  if (non_stop)
    target_dcache_invalidate ();



It seems like we can "dangerously" hit stale cache (probably most
visible with non-stop mode) between target events:

 #1 - all threads set running, dcache is invalidated

 #2 - thread 1 stops.  As we handle the event, we read code, and cache it.

 #3 - others threads continue running.  some thread jits something, or
   changes code that was cached.

 #4 - thread 2 stops.  As we handle the event, we read code, hitting
   stale cache.


I'm thinking we might need to flush the dcache before handling each
event, like we already invalidate the overlay cache (see
"overlay_cache_invalid = 1" in infrun.c) ?

-- 
Pedro Alves