Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 26 May 2023 12:26:48 -0400
From:      Naman Sood <naman@freebsdfoundation.org>
To:        Mark Johnston <markj@freebsd.org>
Cc:        freebsd-hackers@freebsd.org
Subject:   Re: APIC interrupting me while stepping through a kernel with kgdb
Message-ID:  <CACc-QYctcRUdQOGQf%2BN0uYAR93DNZ5vGA%2B%2BZReaFrMJvT3hgqw@mail.gmail.com>
In-Reply-To: <CACc-QYe8dQLhkVfUDsgkKK2U9wi8ffwt0k-HvT2Ly6OFqLg0rA@mail.gmail.com>
References:  <CACc-QYeCa-hRd7AO9Tpjg7Mt=mReaHDKazJQrX22r7UAyTbxGQ@mail.gmail.com> <ZG_ErodCVMgJsTzN@nuc> <CACc-QYe8dQLhkVfUDsgkKK2U9wi8ffwt0k-HvT2Ly6OFqLg0rA@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Okay, this seems to mostly work, but I am still getting timer
interrupts sometimes. I'll post here if I can get that to be
reproducible, but until then, I can live with this. Thanks!

Naman.

On Thu, May 25, 2023 at 4:33=E2=80=AFPM Naman Sood <naman@freebsdfoundation=
.org> wrote:
>
> Oh. How convenient! I'm running 14.0-CURRENT, though I may have
> delayed system updates long enough to not have this patch at the
> moment. Let me get back to you after I run updates and retry - thanks!
>
> Naman.
>
> On Thu, May 25, 2023 at 4:27=E2=80=AFPM Mark Johnston <markj@freebsd.org>=
 wrote:
> >
> > On Thu, May 25, 2023 at 03:34:56PM -0400, Naman Sood wrote:
> > > Hello!
> > >
> > > I've been following this guide[1] to get kgdb to debug a FreeBSD gues=
t
> > > on bhyve - I have a WIP patch for pfsync that I want to debug. I'm at
> > > the point where I can set breakpoints in if_pfsync.c and get them to
> > > break when I hit them. However, when I step/next from there, control
> > > always goes to the lapic_handle_timer(), which does its APIC timer
> > > things, returns control to the line where I set the breakpoint... and
> > > then immediately goes to the timer again, since presumably a timer
> > > tick has happened between me pressing Return repeatedly.
> > >
> > > Is there a way to get around this? Either turn off the timer (doesn't
> > > sound like a good idea?) or get kgdb to ignore it so I can debug the
> > > rest of the kernel? Running skip at the timer code didn't seem to
> > > help.
> >
> > Which revision of FreeBSD are you running?  I believe this recent commi=
t
> > will address the problem you're describing, but you won't have it unles=
s
> > you're running the main branch:
> > https://cgit.freebsd.org/src/commit/?id=3Dfefac543590d



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CACc-QYctcRUdQOGQf%2BN0uYAR93DNZ5vGA%2B%2BZReaFrMJvT3hgqw>