[prev in list] [next in list] [prev in thread] [next in thread] 

List:       gdb
Subject:    Re: Target sim crashes when recording for reverse debugging; find_inferior_ptid returns null
From:       William Tambe via Gdb <gdb () sourceware ! org>
Date:       2020-05-24 14:00:01
Message-ID: CAF8i9mOUMSejMSf6wMEimf++QA-8one_+TEwpOy-9Fhj9vwP3g () mail ! gmail ! com
[Download RAW message or body]

On Sun, May 24, 2020 at 6:40 AM Andrew Burgess
<andrew.burgess@embecosm.com> wrote:
>
> * William Tambe <tambewilliam@gmail.com> [2020-05-17 12:49:21 -0400]:
>
> > On Sun, May 17, 2020 at 10:09 AM William Tambe <tambewilliam@gmail.com> wrote:
> > >
> > > On Sun, May 17, 2020 at 4:12 AM Andrew Burgess
> > > <andrew.burgess@embecosm.com> wrote:
> > > >
> > > > * William Tambe via Gdb <gdb@sourceware.org> [2020-05-16 08:56:20 -0400]:
> > > >
> > > > > ../../iron-toolchain/binutils/gdb/process-stratum-target.c:47:
> > > > > internal-error: virtual gdbarch*
> > > > > process_stratum_target::thread_architecture(ptid_t): Assertion `inf !=
> > > > > NULL' failed.
> > > > >
> > > > > Above is the crash that occur after issuing following gdb commands:
> > > > > target sim
> > > > > load
> > > > > record
> > > > > si
> > > > >
> > > > > The crash occurs because find_inferior_ptid() returns NULL.
> > > > > From below backtrace, is it possible to determine what could be the
> > > > > cause of find_inferior_ptid() returning NULL ?
> > > >
> > > > I've been running with the patch below in place for doing reverse
> > > > debugging on a simulator.
> > >
> > > Should I revert this change if I need to use "target exec" ?
> >
> > In other words, is this change ok for both "target sim" and "target
> > exec" ?
>
> It should be fine for all targets.

Great, thanks!

>
> Thanks,
> Andrew
>
>
>
> >
> > >
> > > >
> > > > I need to clean this up and get it merged at some point.
> > > >
> > > > Hope this helps,
> > > > Thanks,
> > > > Andrew
> > > >
> > > >
> > > > ---
> > > >
> > > > commit ecc84c55f926c5a3d7c0dd16b8f46df597ed3f5f
> > > > Author: Andrew Burgess <andrew.burgess@embecosm.com>
> > > > Date:   Mon Oct 21 21:17:00 2019 +0100
> > > >
> > > >     gdb: Avoid target_thread_architecture
> > > >
> > > >     Don't call target_thread_architecture, get gdbarch from regcache
> > > >     instead.
> > > >
> > > > diff --git a/gdb/record-full.c b/gdb/record-full.c
> > > > index 51b7beabf66..8c83b017b26 100644
> > > > --- a/gdb/record-full.c
> > > > +++ b/gdb/record-full.c
> > > > @@ -1074,9 +1074,10 @@ record_full_target::resume (ptid_t ptid, int step, enum gdb_signal signal)
> > > >
> > > >    if (!RECORD_FULL_IS_REPLAY)
> > > >      {
> > > > -      struct gdbarch *gdbarch = target_thread_architecture (ptid);
> > > > +      struct regcache *regcache = get_current_regcache ();
> > > > +      struct gdbarch *gdbarch = regcache->arch ();
> > > >
> > > > -      record_full_message (get_current_regcache (), signal);
> > > > +      record_full_message (regcache, signal);
> > > >
> > > >        if (!step)
> > > >          {
> > > >
> > > >
> > > >
[prev in list] [next in list] [prev in thread] [next in thread]