Re: kern/56561: cv_is_valid assertion failure in intel drm

[Patrick Welche <prlw1%cam.ac.uk@localhost>]

The following reply was made to PR kern/56561; it has been noted by GNATS.

From: Patrick Welche <prlw1%cam.ac.uk@localhost>
To: Taylor R Campbell <riastradh%NetBSD.org@localhost>
Cc: gnats-bugs%NetBSD.org@localhost
Subject: Re: kern/56561: cv_is_valid assertion failure in intel drm
Date: Sun, 26 Dec 2021 18:11:16 +0000

 On Sun, Dec 26, 2021 at 05:57:50PM +0000, Taylor R Campbell wrote:
 > > Date: Sun, 26 Dec 2021 17:54:13 +0000
 > > From: Patrick Welche <prlw1%cam.ac.uk@localhost>
 > > 
 > > Indeed it is! It turns out the trap happens more frequently, and
 > > usually when idle. This is with a 22 Dec 9.99.93 kernel:
 > > 
 > > prevented execution of 0x0 (SMEP)
 > > [...]
 > > intel_disable_ddi() at netbsd:intel_disable_ddi+0xa3
 > 
 > Got a line number for intel_disable_ddi+0xa3?
 
 Sort of? A3=163
 
 Dump of assembler code for function intel_disable_ddi:
    0xffffffff804e9d6d <+0>:       push   %rbp
 ...
    0xffffffff804e9e0b <+158>:     call   0xffffffff80521597 <intel_edp_backlight_off>
    0xffffffff804e9e10 <+163>:     xor    %edx,%edx    <----
    0xffffffff804e9e12 <+165>:     mov    %r12,%rsi
    0xffffffff804e9e15 <+168>:     mov    %r15,%rdi
    0xffffffff804e9e18 <+171>:     pop    %r12
    0xffffffff804e9e1a <+173>:     pop    %r13
    0xffffffff804e9e1c <+175>:     pop    %r14
    0xffffffff804e9e1e <+177>:     pop    %r15
    0xffffffff804e9e20 <+179>:     pop    %rbp
    0xffffffff804e9e21 <+180>:     jmp    0xffffffff805215fd <intel_dp_sink_set_decompression_state>
 
 (I'm clearly net getting addr2line right...)