Re: [PATCH] Fix unwind_frame for clang-built kernels

2020-06-18 Thread Miles Chen
On Thu, 2020-06-18 at 13:22 -0700, Nick Desaulniers wrote: > On Tue, Jun 16, 2020 at 3:36 PM 'Nathan Huckleberry' via Clang Built > Linux wrote: > > > > Since clang does not push pc and sp in function prologues, the current > > implementation of unwind_frame does not work. By using the previous >

Re: [PATCH] Fix unwind_frame for clang-built kernels

2020-06-18 Thread Russell King - ARM Linux admin
On Wed, Jun 17, 2020 at 06:23:48AM +0200, Sedat Dilek wrote: > On Wed, Jun 17, 2020 at 12:36 AM 'Nathan Huckleberry' via Clang Built > Linux wrote: > > > > Since clang does not push pc and sp in function prologues, the current > > implementation of unwind_frame does not work. By using the previous

Re: [PATCH] Fix unwind_frame for clang-built kernels

2020-06-18 Thread Nick Desaulniers
On Thu, Jun 18, 2020 at 1:22 PM Nick Desaulniers wrote: > > On Tue, Jun 16, 2020 at 3:36 PM 'Nathan Huckleberry' via Clang Built > Linux wrote: > > > > Since clang does not push pc and sp in function prologues, the current > > implementation of unwind_frame does not work. By using the previous >

Re: [PATCH] Fix unwind_frame for clang-built kernels

2020-06-18 Thread Nick Desaulniers
On Tue, Jun 16, 2020 at 3:36 PM 'Nathan Huckleberry' via Clang Built Linux wrote: > > Since clang does not push pc and sp in function prologues, the current > implementation of unwind_frame does not work. By using the previous > frame's lr/fp instead of saved pc/sp we get valid unwinds on clang-bu

Re: [PATCH] Fix unwind_frame for clang-built kernels

2020-06-16 Thread Sedat Dilek
On Wed, Jun 17, 2020 at 12:36 AM 'Nathan Huckleberry' via Clang Built Linux wrote: > > Since clang does not push pc and sp in function prologues, the current > implementation of unwind_frame does not work. By using the previous > frame's lr/fp instead of saved pc/sp we get valid unwinds on clang-b

[PATCH] Fix unwind_frame for clang-built kernels

2020-06-16 Thread Nathan Huckleberry
Since clang does not push pc and sp in function prologues, the current implementation of unwind_frame does not work. By using the previous frame's lr/fp instead of saved pc/sp we get valid unwinds on clang-built kernels. The bounds check on next frame pointer must be changed as well since there ar