I have given a respin to the tb_lock branch on top of master.
Turns out the freezes I was getting were due to the magic we used
to rely on (e.g. for icount). Works now!
The branch is here:
https://github.com/cota/qemu/tree/tb-lock
Boot-tested on ppc64, aarch64 and alpha smp guests (<=64) with M
On Fri, Oct 20, 2017 at 09:10:38 +0200, Paolo Bonzini wrote:
> >> 2) in tb_for_each_tagged_safe, could the "prev" argument instead be
> >> "next", like
> >
> > Is this just to make them closer to the macros in queue.h?
> >
> > In this case tracking *prev in the loop (rather than next) is
> > usef
On 19/10/2017 22:11, Emilio G. Cota wrote:
> On Thu, Oct 19, 2017 at 15:05:17 +0200, Paolo Bonzini wrote:
>> On 19/10/2017 00:45, Emilio G. Cota wrote:
>>> I have just pushed a branch on top of this series that includes
>>> 10 patches that further pave the way for the removal of tb_lock:
>>>
>>>
On Thu, Oct 19, 2017 at 15:05:17 +0200, Paolo Bonzini wrote:
> On 19/10/2017 00:45, Emilio G. Cota wrote:
> > I have just pushed a branch on top of this series that includes
> > 10 patches that further pave the way for the removal of tb_lock:
> >
> > https://github.com/cota/qemu/tree/multi-tcg-v
On 19/10/2017 00:45, Emilio G. Cota wrote:
> On Mon, Oct 16, 2017 at 10:25:19 -0700, Richard Henderson wrote:
>> I've fixed two bugs within v5 of Emilio's patch set:
>>
>> - The step_next_tb patch fixes the "rep movsb" bug that appeared
>>when we included CF_COUNT_MASK into CF_HASH_MASK. We h
On Mon, Oct 16, 2017 at 10:25:19 -0700, Richard Henderson wrote:
> I've fixed two bugs within v5 of Emilio's patch set:
>
> - The step_next_tb patch fixes the "rep movsb" bug that appeared
>when we included CF_COUNT_MASK into CF_HASH_MASK. We had been
>relying on magic to single-step the
On Mon, Oct 16, 2017 at 10:25:19 -0700, Richard Henderson wrote:
>Thus, an extra 20 patches in the patch set rearranging how temps
>are referenced within the TCG backend. Some of them have appeared
>on the list before, but it would have been last November.
I remember seeing these on t
Hi,
This series seems to have some coding style problems. See output below for
more information:
Type: series
Message-id: 20171016172609.23422-1-richard.hender...@linaro.org
Subject: [Qemu-devel] [PATCH v6 00/50] tcg tb_lock removal
=== TEST SCRIPT BEGIN ===
#!/bin/bash
BASE=base
n=1
total
I've fixed two bugs within v5 of Emilio's patch set:
- The step_next_tb patch fixes the "rep movsb" bug that appeared
when we included CF_COUNT_MASK into CF_HASH_MASK. We had been
relying on magic to single-step the next guest insn.
- The original "allocate optimizer temps with tcg_mallo