On 3 January 2013 17:37, Alexander Graf <ag...@suse.de> wrote:
> On 03.01.2013, at 18:28, Peter Maydell wrote:
>> Untested, and I haven't looked up the PPC ABI to check that the
>> function is overall doing the right thing, but with those caveats:
>> Reviewed-by: Peter Maydell <peter.mayd...@linaro.org>
>
> Considering the state it was in before, I'd say applying the patch
> is an improvement regardless on how broken any code around it
> might be :).
Forgot, this patch doesn't address the other issue I mentioned,
where
env->gpr[4] = (target_ulong) h2g(sc);
is passing the guest a pointer to potentially about to be freed
memory and should be doing something like
env->gpr[4] = frame_addr + offsetof(struct target_sigframe, sctx);
instead.
-- PMM
