Re: port-sparc/54734 (sparc panics running ATF tests)

[Kamil Rytarowski <n54%gmx.com@localhost>]

The following reply was made to PR port-sparc/54734; it has been noted by GNATS.

From: Kamil Rytarowski <n54%gmx.com@localhost>
To: Martin Husemann <martin%duskware.de@localhost>,
 "gnats-bugs%NetBSD.org@localhost" <gnats-bugs%NetBSD.org@localhost>
Cc: 
Subject: Re: port-sparc/54734 (sparc panics running ATF tests)
Date: Sat, 25 Jan 2020 18:40:33 +0100

 On 25.01.2020 10:36, Martin Husemann wrote:
 > On Fri, Jan 24, 2020 at 10:05:01PM +0000, Kamil Rytarowski wrote:
 >>  http://netbsd.org/~kamil/patch-00216-sparc-ptrace-panic.txt
 >
 > That change is fine, is it enough to fix that panic?
 >
 
 It should be everything.
 
 > Two notes:
 >
 >  - we need the same in sparc64/sparc64/process_machdep.c (not sure why
 >    that didn't show up in test runs so far)
 
 In theory every alignment sensitive port is endangered.
 
 https://github.com/openbsd/src/blob/master/regress/sys/ptrace/ptrace.c
 
 Context:
 
 https://marc.info/?l=3Dopenbsd-bugs&m=3D107558043319084&w=3D2
 
 In theory just using PTRACE_REG_SET_PC() might not be enough to trigger
 a panic on a certain port.
 
 >  - in process_write_regs() we should also verify that npc !=3D pc
 >    (or force npc to be pc + 4 if it isn't)
 >
 
 If there is just triggered a crash signal that kills the child process
 but not the kernel, we could ignore it in the kernel.
 
 > Martin
 >