Re: [PATCH 0/4 V2] gdbarch software_single_step returns VEC (CORE_ADDR) *

[Yao Qi <qiyaoltc at gmail dot com>]

Yao Qi <qiyaoltc@gmail.com> writes:

Hi Pedro,
Your comments to V1 are addressed in V2.  Can you take a look?

> This is a step toward changing gdbarch software_single_step to
> "F:VEC (CORE_ADDR) *:software_single_step:struct regcache *regcache:regcache"
> as I mentioned https://sourceware.org/ml/gdb-patches/2015-11/msg00591.html
>
> It isn't trivial to replace frame_info with regcache because it may
> break the target that frames have different gdbarch, such as spu, so
> I still think about it.  In the mean time, I post this series, to let
> software_single_step returns VEC (CORE_ADDR) *, and
> displaced_step_hw_singlestep can be removed.
>
> V1 can be found https://sourceware.org/ml/gdb-patches/2016-03/msg00466.html
> The main change, compared with V1, is that we finally remove the global
> variable arm_override_mode.  During the review to V1, there are two
> suggested ways, but they need big changes on gdb target/gdbarch
> interfaces.  Finally, I get a simpler change in patch 1.  The rest of
> patches are straightforward.
>
> I don't include patch "Remove gdbarch method displaced_step_hw_singlestep"
> in this series, because it causes regression on aarch64-linux doing
> multi-arch debugging, and I am looking at it.
>
> Regression tested on arm-linux, x86_64-linux and aarch64-linux.

-- 
Yao (éå)