Messages in this thread | | | Date | Fri, 22 May 2020 17:35:08 +0100 | From | Daniel Thompson <> | Subject | Re: [PATCH v2] kdb: Make kdb_printf robust to run in NMI context |
| |
On Fri, May 22, 2020 at 08:03:47PM +0530, Sumit Garg wrote: > While rounding up CPUs via NMIs, its possible that a rounded up CPU > maybe holding a console port lock leading to kgdb master CPU stuck in > a deadlock during invocation of console write operations. So in order > to avoid such a deadlock, invoke bust_spinlocks() prior to invocation > of console handlers. > > Also, add a check for console port to be enabled prior to invocation of > corresponding handler.
Perhaps this should have been two patches.
In fact, to be honest, I'd suggest combining all the patches to improve kdb console handling (including a fixed version of the RFC) into a single patch set.
> Suggested-by: Petr Mladek <pmladek@suse.com> > Suggested-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com> > Signed-off-by: Sumit Garg <sumit.garg@linaro.org> > --- > > Changes in v2: > - Use oops_in_progress directly instead of bust_spinlocks(). > > kernel/debug/kdb/kdb_io.c | 8 ++++++++ > 1 file changed, 8 insertions(+) > > diff --git a/kernel/debug/kdb/kdb_io.c b/kernel/debug/kdb/kdb_io.c > index 924bc92..3a5a068 100644 > --- a/kernel/debug/kdb/kdb_io.c > +++ b/kernel/debug/kdb/kdb_io.c > @@ -699,7 +699,11 @@ int vkdb_printf(enum kdb_msgsrc src, const char *fmt, va_list ap) > } > } > for_each_console(c) { > + if (!(c->flags & CON_ENABLED)) > + continue; > + ++oops_in_progress;
Given the subtly of what is going on I think we need some comments in the code on what we are doing and why.
> c->write(c, cp, retlen - (cp - kdb_buffer)); > + --oops_in_progress; > touch_nmi_watchdog(); > } > } > @@ -761,7 +765,11 @@ int vkdb_printf(enum kdb_msgsrc src, const char *fmt, va_list ap) > } > } > for_each_console(c) { > + if (!(c->flags & CON_ENABLED)) > + continue; > + ++oops_in_progress; > c->write(c, moreprompt, strlen(moreprompt)); > + --oops_in_progress; > touch_nmi_watchdog();
As with the other patches maybe the first patch in the set should be factoring out the common code before making changes to it.
Daniel.
| |