Re: [Qemu-devel] [PATCH v3 4/4] tcg: rework tb_invalidated_flag

[Peter Maydell <peter.maydell@linaro.org>]

On 18 April 2016 at 16:05, Sergey Fedorov <serge.fdrv@gmail.com> wrote:
> @@ -507,14 +510,12 @@ int cpu_exec(CPUState *cpu)
>                  }
>                  tb_lock();
>                  tb = tb_find_fast(cpu);
> -                /* Note: we do it here to avoid a gcc bug on Mac OS X when
> -                   doing it in tb_find_slow */
>
> Is this still true? Would it make more sense to push the patching down
> to the gen_code?
>
>
> This comment comes up to the commit:
>
> commit 1538800276aa7228d74f9d00bf275f54dc9e9b43
> Author: bellard <bellard@c046a42c-6fe2-441c-8c8c-71466251a162>
> Date:   Mon Dec 19 01:42:32 2005 +0000
>
>     workaround for gcc bug on PowerPC

We no longer support building on PPC OSX hosts, and haven't
for some time -- the earliest OSX we support is 10.5, which is
x86-only. So I think we can probably safely dump the gcc bug
workaround, as long as we're clear in commit messages that we're
doing so.

Looking at that commit, it seems to be related to changing the
value of T0 inside the tb_find_slow() function; at that point
in QEMU's history, T0 was a variable which was tied to a specific
host CPU register via gcc's 'asm("registername")' syntax; it's
highly likely that the gcc bug was specific to handling of that
kind of register variable anyway.

thanks
-- PMM