On Mon, 2021-09-20 at 20:32 +0100, Andrew Cooper wrote:
> (Almost) no functional change.
> 
> irq_move_cleanup_interrupt() changes two smp_processor_id() calls to
> the 'me'
> local variable which manifests as a minor code improvement.  All
> other
> differences in the compiled binary are to do with line numbers
> changing.
> 
> Some conversion notes:
>  * HVMTRACE_LONG_[234]D() and TRACE_2_LONG_[234]D() were latently
> buggy.  They
>    blindly discard extra parameters, but luckily no users are
> impacted.  They
>    are also obfuscated wrappers, depending on exactly one or two
> parameters
>    being TRC_PAR_LONG() to compile successfully.
>  * HVMTRACE_LONG_1D() behaves unlike its named companions, and takes
> exactly
>    one 64bit parameter which it splits manually.  It's one user,
>    vmx_cr_access()'s LMSW path, is gets adjusted to use
> TRACE_PARAM64().
>  * TRACE_?D() and TRACE_2_LONG_*() change to TRACE_TIME() as cycles
> is always.
>
Was this supposed to be "as cycles is always 1", or something like
that? (But maybe it's fine and it is me. I'm no native speaker after
all...)

In any case...

>  * HVMTRACE_ND() is opencoded for VMENTRY/VMEXIT records to include
> cycles.
>    These are converted to TRACE_TIME(), with the old modifier
> parameter
>    expressed as an OR at the callsite.  One callsite,
> svm_vmenter_helper() had
>    a nested tb_init_done check, which is dropped.  (The optimiser
> also spotted
>    this, which is why it doesn't manifest as a binary difference.)
>  * All HVMTRACE_?D() change to TRACE() as cycles is explicitly
> skipped.
> 
> Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
>
Reviewed-by: Dario Faggioli <dfaggioli@suse.com>

Regards
-- 
Dario Faggioli, Ph.D
http://about.me/dario.faggioli
Virtualization Software Engineer
SUSE Labs, SUSE https://www.suse.com/
-------------------------------------------------------------------
<<This happens because _I_ choose it to happen!>> (Raistlin Majere)