* [REGRESSION 5.15.y][BISECTED][PATCH 0/1] console hangs at uevents
@ 2025-09-10 9:38 Brett A C Sheffield
2025-09-10 9:38 ` [PATCH 1/1] Revert "fbdev: Disable sysfb device registration when removing conflicting FBs" Brett A C Sheffield
0 siblings, 1 reply; 6+ messages in thread
From: Brett A C Sheffield @ 2025-09-10 9:38 UTC (permalink / raw)
To: stable
Cc: regressions, linux-kernel, linux-fbdev, dri-devel,
Javier Martinez Canillas, Simona Vetter, Helge Deller,
Thomas Zimmermann, Lee Jones, Murad Masimov, Yongzhen Zhang,
Greg Kroah-Hartman, Sasha Levin, Brett A C Sheffield
Commit ee7a69aa38d8 ("fbdev: Disable sysfb device registration when removing
conflicting FBs") was backported to 5.15.y LTS. This causes a regression where
all virtual consoles stop responding during boot at:
"Populating /dev with existing devices through uevents ..."
The console is no longer responding, nor is it possible to switch to another
virtual console. It is possible to force a reboot with CTRL+ALT+DEL.
The system is otherwise still working normally, and can be connected to over
ssh.
This patch was not backported to 5.4.y or 5.10.y. Newer stable kernels with this
patch (6.1.y, 6.6.y, 6.12,y, 6.15.y, 6.16.y) and mainline are unaffected.
There are a bunch of patches that touch fbmem.c in later stable kernels, and I
started looking at this but it gets messy very quickly as e23a5e14aa27 is a
merge commit.
Given this was an autoselected patch that was not backported to fix any specific
issue the simplest and cleanest fix is to simply revert.
Bisection log:
git bisect start
# status: waiting for both good and bad commits
# good: [8bb7eca972ad531c9b149c0a51ab43a417385813] Linux 5.15
git bisect good 8bb7eca972ad531c9b149c0a51ab43a417385813
# status: waiting for bad commit, 1 good commit known
# bad: [c275eaaaa34260e6c907bc5e7ee07c096bc45064] Linux 5.15.125
git bisect bad c275eaaaa34260e6c907bc5e7ee07c096bc45064
# bad: [ff2932ac8ee1ce6f66ba1b5017843c87492cd1a8] tcp: Fix a data-race around sysctl_tcp_challenge_ack_limit.
git bisect bad ff2932ac8ee1ce6f66ba1b5017843c87492cd1a8
# good: [72ea28d88d30cefc5b0184e4666cf4817f2a0fa8] arm64: dts: marvell: armada-37xx: Remap IO space to bus address 0x0
git bisect good 72ea28d88d30cefc5b0184e4666cf4817f2a0fa8
# good: [acd12d16528152b32fa09be2c5ef95047f69af05] KVM: x86/mmu: fix NULL pointer dereference on guest INVPCID
git bisect good acd12d16528152b32fa09be2c5ef95047f69af05
# good: [ce49b94ddb70e607c476a12d26002715d694b23a] random: avoid checking crng_ready() twice in random_init()
git bisect good ce49b94ddb70e607c476a12d26002715d694b23a
# good: [6886327780254ba749b770373653b6afc2a339fc] fbcon: Disallow setting font bigger than screen size
git bisect good 6886327780254ba749b770373653b6afc2a339fc
# bad: [1f068f9da7436b11276f23ebf65dd641c8ae0b62] x86/entry: Avoid very early RET
git bisect bad 1f068f9da7436b11276f23ebf65dd641c8ae0b62
# good: [5a4bb158f4c638d1fc494d2730357cfb6f5a4eea] netfilter: nf_tables: replace BUG_ON by element length check
git bisect good 5a4bb158f4c638d1fc494d2730357cfb6f5a4eea
# bad: [e71d0e1370b162271a0053e1f1215961c31dc4cb] ASoC: codecs: rt700/rt711/rt711-sdca: initialize workqueues in probe
git bisect bad e71d0e1370b162271a0053e1f1215961c31dc4cb
# good: [2d84fcb6e6f74c96eb03e6a1b9eed0a164e37b2c] seg6: bpf: fix skb checksum in bpf_push_seg6_encap()
git bisect good 2d84fcb6e6f74c96eb03e6a1b9eed0a164e37b2c
# good: [06f818de1621ea4da689fd45f1eb54ff7617342b] scsi: hisi_sas: Limit max hw sectors for v3 HW
git bisect good 06f818de1621ea4da689fd45f1eb54ff7617342b
# bad: [7a2294c5f2e5636772afe6bd6c5b28218e0ea154] nvme: fix regression when disconnect a recovering ctrl
git bisect bad 7a2294c5f2e5636772afe6bd6c5b28218e0ea154
# good: [b952569e03168e23258fabf3faa4d6d2b8e08835] firmware: sysfb: Add sysfb_disable() helper function
git bisect good b952569e03168e23258fabf3faa4d6d2b8e08835
# bad: [833ecd0eae76eadf81d6d747bb5bc992d1151867] net: tipc: fix possible refcount leak in tipc_sk_create()
git bisect bad 833ecd0eae76eadf81d6d747bb5bc992d1151867
# bad: [13d28e0c79cbf69fc6f145767af66905586c1249] fbdev: Disable sysfb device registration when removing conflicting FBs
git bisect bad 13d28e0c79cbf69fc6f145767af66905586c1249
# first bad commit: [13d28e0c79cbf69fc6f145767af66905586c1249] fbdev: Disable sysfb device registration when removing conflicting FBs
Brett A C Sheffield (1):
Revert "fbdev: Disable sysfb device registration when removing
conflicting FBs"
drivers/video/fbdev/core/fbmem.c | 12 ------------
1 file changed, 12 deletions(-)
--
2.49.1
^ permalink raw reply [flat|nested] 6+ messages in thread* [PATCH 1/1] Revert "fbdev: Disable sysfb device registration when removing conflicting FBs"
2025-09-10 9:38 [REGRESSION 5.15.y][BISECTED][PATCH 0/1] console hangs at uevents Brett A C Sheffield
@ 2025-09-10 9:38 ` Brett A C Sheffield
2025-09-10 10:46 ` Javier Martinez Canillas
0 siblings, 1 reply; 6+ messages in thread
From: Brett A C Sheffield @ 2025-09-10 9:38 UTC (permalink / raw)
To: stable
Cc: regressions, linux-kernel, linux-fbdev, dri-devel,
Javier Martinez Canillas, Simona Vetter, Helge Deller,
Thomas Zimmermann, Lee Jones, Murad Masimov, Yongzhen Zhang,
Greg Kroah-Hartman, Sasha Levin, Brett A C Sheffield
This reverts commit 13d28e0c79cbf69fc6f145767af66905586c1249.
Commit ee7a69aa38d8 ("fbdev: Disable sysfb device registration when
removing conflicting FBs") was backported to 5.15.y LTS. This causes a
regression where all virtual consoles stop responding during boot at:
"Populating /dev with existing devices through uevents ..."
Reverting the commit fixes the regression.
Signed-off-by: Brett A C Sheffield <bacs@librecast.net>
---
drivers/video/fbdev/core/fbmem.c | 12 ------------
1 file changed, 12 deletions(-)
diff --git a/drivers/video/fbdev/core/fbmem.c b/drivers/video/fbdev/core/fbmem.c
index d938c31e8f90..3b52ddfe0350 100644
--- a/drivers/video/fbdev/core/fbmem.c
+++ b/drivers/video/fbdev/core/fbmem.c
@@ -19,7 +19,6 @@
#include <linux/kernel.h>
#include <linux/major.h>
#include <linux/slab.h>
-#include <linux/sysfb.h>
#include <linux/mm.h>
#include <linux/mman.h>
#include <linux/vt.h>
@@ -1795,17 +1794,6 @@ int remove_conflicting_framebuffers(struct apertures_struct *a,
do_free = true;
}
- /*
- * If a driver asked to unregister a platform device registered by
- * sysfb, then can be assumed that this is a driver for a display
- * that is set up by the system firmware and has a generic driver.
- *
- * Drivers for devices that don't have a generic driver will never
- * ask for this, so let's assume that a real driver for the display
- * was already probed and prevent sysfb to register devices later.
- */
- sysfb_disable();
-
mutex_lock(®istration_lock);
do_remove_conflicting_framebuffers(a, name, primary);
mutex_unlock(®istration_lock);
--
2.49.1
^ permalink raw reply related [flat|nested] 6+ messages in thread* Re: [PATCH 1/1] Revert "fbdev: Disable sysfb device registration when removing conflicting FBs"
2025-09-10 9:38 ` [PATCH 1/1] Revert "fbdev: Disable sysfb device registration when removing conflicting FBs" Brett A C Sheffield
@ 2025-09-10 10:46 ` Javier Martinez Canillas
2025-09-10 10:52 ` Brett A C Sheffield
0 siblings, 1 reply; 6+ messages in thread
From: Javier Martinez Canillas @ 2025-09-10 10:46 UTC (permalink / raw)
To: Brett A C Sheffield, stable
Cc: regressions, linux-kernel, linux-fbdev, dri-devel, Simona Vetter,
Helge Deller, Thomas Zimmermann, Lee Jones, Murad Masimov,
Yongzhen Zhang, Greg Kroah-Hartman, Sasha Levin,
Brett A C Sheffield
Brett A C Sheffield <bacs@librecast.net> writes:
Hello Brett,
> This reverts commit 13d28e0c79cbf69fc6f145767af66905586c1249.
>
> Commit ee7a69aa38d8 ("fbdev: Disable sysfb device registration when
> removing conflicting FBs") was backported to 5.15.y LTS. This causes a
> regression where all virtual consoles stop responding during boot at:
>
> "Populating /dev with existing devices through uevents ..."
>
> Reverting the commit fixes the regression.
>
> Signed-off-by: Brett A C Sheffield <bacs@librecast.net>
> ---
In the other email you said:
> Newer stable kernels with this
> patch (6.1.y, 6.6.y, 6.12,y, 6.15.y, 6.16.y) and mainline are unaffected.
But are you proposing to revert the mentioned commit in mainline too
or just in the 5.15.y LTS tree ?
--
Best regards,
Javier Martinez Canillas
Core Platforms
Red Hat
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: [PATCH 1/1] Revert "fbdev: Disable sysfb device registration when removing conflicting FBs"
2025-09-10 10:46 ` Javier Martinez Canillas
@ 2025-09-10 10:52 ` Brett A C Sheffield
2025-09-10 11:01 ` Sasha Levin
2025-09-10 11:09 ` Javier Martinez Canillas
0 siblings, 2 replies; 6+ messages in thread
From: Brett A C Sheffield @ 2025-09-10 10:52 UTC (permalink / raw)
To: Javier Martinez Canillas
Cc: stable, regressions, linux-kernel, linux-fbdev, dri-devel,
Simona Vetter, Helge Deller, Thomas Zimmermann, Lee Jones,
Murad Masimov, Yongzhen Zhang, Greg Kroah-Hartman, Sasha Levin
On 2025-09-10 12:46, Javier Martinez Canillas wrote:
> Brett A C Sheffield <bacs@librecast.net> writes:
>
> Hello Brett,
>
> > This reverts commit 13d28e0c79cbf69fc6f145767af66905586c1249.
> >
> > Commit ee7a69aa38d8 ("fbdev: Disable sysfb device registration when
> > removing conflicting FBs") was backported to 5.15.y LTS. This causes a
> > regression where all virtual consoles stop responding during boot at:
> >
> > "Populating /dev with existing devices through uevents ..."
> >
> > Reverting the commit fixes the regression.
> >
> > Signed-off-by: Brett A C Sheffield <bacs@librecast.net>
> > ---
>
> In the other email you said:
>
> > Newer stable kernels with this
> > patch (6.1.y, 6.6.y, 6.12,y, 6.15.y, 6.16.y) and mainline are unaffected.
>
> But are you proposing to revert the mentioned commit in mainline too
> or just in the 5.15.y LTS tree ?
Only the 5.15.y tree. Sorry - that could have been clearer. There's no
regression anywhere else. Mainline and other stable kernels are all ok.
Cheers,
Brett
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: [PATCH 1/1] Revert "fbdev: Disable sysfb device registration when removing conflicting FBs"
2025-09-10 10:52 ` Brett A C Sheffield
@ 2025-09-10 11:01 ` Sasha Levin
2025-09-10 11:09 ` Javier Martinez Canillas
1 sibling, 0 replies; 6+ messages in thread
From: Sasha Levin @ 2025-09-10 11:01 UTC (permalink / raw)
To: Brett A C Sheffield
Cc: Javier Martinez Canillas, stable, regressions, linux-kernel,
linux-fbdev, dri-devel, Simona Vetter, Helge Deller,
Thomas Zimmermann, Lee Jones, Murad Masimov, Yongzhen Zhang,
Greg Kroah-Hartman
On Wed, Sep 10, 2025 at 12:52:41PM +0200, Brett A C Sheffield wrote:
>On 2025-09-10 12:46, Javier Martinez Canillas wrote:
>> Brett A C Sheffield <bacs@librecast.net> writes:
>>
>> Hello Brett,
>>
>> > This reverts commit 13d28e0c79cbf69fc6f145767af66905586c1249.
>> >
>> > Commit ee7a69aa38d8 ("fbdev: Disable sysfb device registration when
>> > removing conflicting FBs") was backported to 5.15.y LTS. This causes a
>> > regression where all virtual consoles stop responding during boot at:
>> >
>> > "Populating /dev with existing devices through uevents ..."
>> >
>> > Reverting the commit fixes the regression.
>> >
>> > Signed-off-by: Brett A C Sheffield <bacs@librecast.net>
>> > ---
>>
>> In the other email you said:
>>
>> > Newer stable kernels with this
>> > patch (6.1.y, 6.6.y, 6.12,y, 6.15.y, 6.16.y) and mainline are unaffected.
>>
>> But are you proposing to revert the mentioned commit in mainline too
>> or just in the 5.15.y LTS tree ?
>
>Only the 5.15.y tree. Sorry - that could have been clearer. There's no
>regression anywhere else. Mainline and other stable kernels are all ok.
Thanks for investigating this! I'll queue it up for 5.15.
--
Thanks,
Sasha
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: [PATCH 1/1] Revert "fbdev: Disable sysfb device registration when removing conflicting FBs"
2025-09-10 10:52 ` Brett A C Sheffield
2025-09-10 11:01 ` Sasha Levin
@ 2025-09-10 11:09 ` Javier Martinez Canillas
1 sibling, 0 replies; 6+ messages in thread
From: Javier Martinez Canillas @ 2025-09-10 11:09 UTC (permalink / raw)
To: Brett A C Sheffield
Cc: stable, regressions, linux-kernel, linux-fbdev, dri-devel,
Simona Vetter, Helge Deller, Thomas Zimmermann, Lee Jones,
Murad Masimov, Yongzhen Zhang, Greg Kroah-Hartman, Sasha Levin
Brett A C Sheffield <bacs@librecast.net> writes:
> On 2025-09-10 12:46, Javier Martinez Canillas wrote:
>> Brett A C Sheffield <bacs@librecast.net> writes:
>>
>> Hello Brett,
>>
>> > This reverts commit 13d28e0c79cbf69fc6f145767af66905586c1249.
>> >
>> > Commit ee7a69aa38d8 ("fbdev: Disable sysfb device registration when
>> > removing conflicting FBs") was backported to 5.15.y LTS. This causes a
>> > regression where all virtual consoles stop responding during boot at:
>> >
>> > "Populating /dev with existing devices through uevents ..."
>> >
>> > Reverting the commit fixes the regression.
>> >
>> > Signed-off-by: Brett A C Sheffield <bacs@librecast.net>
>> > ---
>>
>> In the other email you said:
>>
>> > Newer stable kernels with this
>> > patch (6.1.y, 6.6.y, 6.12,y, 6.15.y, 6.16.y) and mainline are unaffected.
>>
>> But are you proposing to revert the mentioned commit in mainline too
>> or just in the 5.15.y LTS tree ?
>
> Only the 5.15.y tree. Sorry - that could have been clearer. There's no
> regression anywhere else. Mainline and other stable kernels are all ok.
>
That's what I thought but just wanted to confirm that was the case. Thanks!
> Cheers,
>
>
> Brett
>
--
Best regards,
Javier Martinez Canillas
Core Platforms
Red Hat
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2025-09-10 11:09 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-09-10 9:38 [REGRESSION 5.15.y][BISECTED][PATCH 0/1] console hangs at uevents Brett A C Sheffield
2025-09-10 9:38 ` [PATCH 1/1] Revert "fbdev: Disable sysfb device registration when removing conflicting FBs" Brett A C Sheffield
2025-09-10 10:46 ` Javier Martinez Canillas
2025-09-10 10:52 ` Brett A C Sheffield
2025-09-10 11:01 ` Sasha Levin
2025-09-10 11:09 ` Javier Martinez Canillas
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).