Re: [PATCH] x86: Drop KGDB support

[Peter Zijlstra]

On Wed, Oct 28, 2026 at 07:15:46AM -0700, H. Peter Anvin wrote:
> On October 28, 2026 5:21:24 AM PDT, Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
> >
> >KGDB does some horrible things, and with everybody having access to
> >virtualization these days, there is really no reason to use it over
> >Qemu's gdb-server.
> >
> >Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>

<snips a gazillion lines>

> Is that really true? What if the debugging depends on specific hardware?

Use device pass-through, emulate it, or printk() debug it.

Using KGDB can cause a ton of problems, it more or less stays stuck in
NMI context, which can terminally screw up the timekeeping code, it can
also cause the watchdogs to fire and there's endless annoyance with code
patching and breakpoints.

People have been trying to 'fix' the breakpoint code to be less broken
in the face of KGBD, but its all terrible and really not worth it. By
fat the sanest option is to just delete KGDB as proposed here.

Also, when was the last time you used KGDB? I think I've tried it once,
some 20 years ago, and haven't bothered since -- it was *that* good :-)