Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.debian.kernel > #73779 > unrolled thread
| Started by | Simon McVittie <smcv@debian.org> |
|---|---|
| First post | 2021-11-24 12:50 +0100 |
| Last post | 2021-11-26 17:10 +0100 |
| Articles | 6 — 3 participants |
Back to article view | Back to linux.debian.kernel
Bug#1000504: linux-image-5.15.0-1-amd64: FUSE regression in 5.15.3 affecting xdg-desktop-portal Simon McVittie <smcv@debian.org> - 2021-11-24 12:50 +0100
Bug#1000504: linux-image-5.15.0-1-amd64: FUSE regression in 5.15.3 affecting xdg-desktop-portal Salvatore Bonaccorso <carnil@debian.org> - 2021-11-24 21:00 +0100
Bug#1000504: linux-image-5.15.0-1-amd64: FUSE regression in 5.15.3 affecting xdg-desktop-portal Simon McVittie <smcv@debian.org> - 2021-11-26 12:10 +0100
Bug#1000504: linux-image-5.15.0-1-amd64: FUSE regression in 5.15.3 affecting xdg-desktop-portal Salvatore Bonaccorso <carnil@debian.org> - 2021-11-26 14:00 +0100
Bug#1000504: linux-image-5.15.0-1-amd64: FUSE regression in 5.15.3 affecting xdg-desktop-portal Simon McVittie <smcv@debian.org> - 2021-12-06 12:10 +0100
Bug#1000504: marked as done (linux-image-5.15.0-1-amd64: FUSE regression in 5.15.3 affecting xdg-desktop-portal) "Debian Bug Tracking System" <owner@bugs.debian.org> - 2021-11-26 17:10 +0100
| From | Simon McVittie <smcv@debian.org> |
|---|---|
| Date | 2021-11-24 12:50 +0100 |
| Subject | Bug#1000504: linux-image-5.15.0-1-amd64: FUSE regression in 5.15.3 affecting xdg-desktop-portal |
| Message-ID | <DmYN3-7h-3@gated-at.bofh.it> |
Package: src:linux
Version: 5.15.3-1
Severity: normal
Tags: upstream
Control: affects -1 xdg-desktop-portal flatpak
Linux 5.15.3, 5.14.20 and other stable kernels are reported to have a
regression that causes Flatpak apps to freeze when using a FUSE filesystem
to load files from the host system. I can't reproduce this myself, but
it has been seen in at least Red Hat and Gentoo, so I wanted to raise
it in Debian as something to watch out for.
The regression was triggered by:
https://lore.kernel.org/all/20211115165430.669780058@linuxfoundation.org/
"fuse: fix page stealing", 712a951025c0667ff00b25afc360f74e639dfabe upstream
https://lore.kernel.org/all/CAOssrKez1mnF4rWRvWgk4LJ2iDfX8xkpMKvgprFt+-ARs83=nA@mail.gmail.com/
is a possible solution (untested - as I said, I couldn't reproduce this
myself).
Tracked elsewhere:
https://github.com/flatpak/flatpak/issues/4595
https://bugzilla.redhat.com/show_bug.cgi?id=2025285
smcv
[toc] | [next] | [standalone]
| From | Salvatore Bonaccorso <carnil@debian.org> |
|---|---|
| Date | 2021-11-24 21:00 +0100 |
| Message-ID | <Dn6AV-4HG-1@gated-at.bofh.it> |
| In reply to | #73779 |
Hi,
On Wed, Nov 24, 2021 at 11:33:08AM +0000, Simon McVittie wrote:
> Package: src:linux
> Version: 5.15.3-1
> Severity: normal
> Tags: upstream
> Control: affects -1 xdg-desktop-portal flatpak
>
> Linux 5.15.3, 5.14.20 and other stable kernels are reported to have a
> regression that causes Flatpak apps to freeze when using a FUSE filesystem
> to load files from the host system. I can't reproduce this myself, but
> it has been seen in at least Red Hat and Gentoo, so I wanted to raise
> it in Debian as something to watch out for.
>
> The regression was triggered by:
> https://lore.kernel.org/all/20211115165430.669780058@linuxfoundation.org/
> "fuse: fix page stealing", 712a951025c0667ff00b25afc360f74e639dfabe upstream
>
> https://lore.kernel.org/all/CAOssrKez1mnF4rWRvWgk4LJ2iDfX8xkpMKvgprFt+-ARs83=nA@mail.gmail.com/
> is a possible solution (untested - as I said, I couldn't reproduce this
> myself).
>
> Tracked elsewhere:
> https://github.com/flatpak/flatpak/issues/4595
> https://bugzilla.redhat.com/show_bug.cgi?id=2025285
And looks there is now a fix for it:
https://lore.kernel.org/stable/CAOssrKez1mnF4rWRvWgk4LJ2iDfX8xkpMKvgprFt+-ARs83=nA@mail.gmail.com/
If you in meanwhile were able to reproduce it and can test the patch
as well that would be great.
712a951025c0 ("fuse: fix page stealing") in any case was backported as
well to 5.10.80, so would need to make sure that we do not release in
next point release a version newly affected.
Regards,
Salvatore
[toc] | [prev] | [next] | [standalone]
| From | Simon McVittie <smcv@debian.org> |
|---|---|
| Date | 2021-11-26 12:10 +0100 |
| Message-ID | <DnHh7-1EU-3@gated-at.bofh.it> |
| In reply to | #73786 |
On Wed, 24 Nov 2021 at 20:55:06 +0100, Salvatore Bonaccorso wrote:
> And looks there is now a fix for it:
>
> https://lore.kernel.org/stable/CAOssrKez1mnF4rWRvWgk4LJ2iDfX8xkpMKvgprFt+-ARs83=nA@mail.gmail.com/
According to a Fedora kernel maintainer on that thread, that proposed
patch seems to work for multiple people.
> If you in meanwhile were able to reproduce it and can test the patch
> as well that would be great.
I tried again to reproduce the issue; this time I was successful and my
laptop became completely unresponsive (couldn't even reboot with the
magic sysrq key, had to power off by holding the power button).
Steps:
* flatpak run org.gnome.Recipes
* New Recipe
* click the "+" below the placeholder photo
* load some random image (it might have to be one that is of a significant
size, when I couldn't reproduce it I think I was using very small test
images from libsdl2-gfx)
I don't know how reliably reproducible this is. Any Flatpak app that uses
the documents portal and does not have direct filesystem access should be
equally suitable.
My laptop is not particularly suitable for building distro kernels,
but I'll try... please don't wait for me though!
smcv
[toc] | [prev] | [next] | [standalone]
| From | Salvatore Bonaccorso <carnil@debian.org> |
|---|---|
| Date | 2021-11-26 14:00 +0100 |
| Message-ID | <DnIZz-2t7-1@gated-at.bofh.it> |
| In reply to | #73803 |
Hi Simon, On Fri, Nov 26, 2021 at 11:00:24AM +0000, Simon McVittie wrote: > On Wed, 24 Nov 2021 at 20:55:06 +0100, Salvatore Bonaccorso wrote: > > And looks there is now a fix for it: > > > > https://lore.kernel.org/stable/CAOssrKez1mnF4rWRvWgk4LJ2iDfX8xkpMKvgprFt+-ARs83=nA@mail.gmail.com/ > > According to a Fedora kernel maintainer on that thread, that proposed > patch seems to work for multiple people. > > > If you in meanwhile were able to reproduce it and can test the patch > > as well that would be great. > > I tried again to reproduce the issue; this time I was successful and my > laptop became completely unresponsive (couldn't even reboot with the > magic sysrq key, had to power off by holding the power button). > > Steps: > > * flatpak run org.gnome.Recipes > * New Recipe > * click the "+" below the placeholder photo > * load some random image (it might have to be one that is of a significant > size, when I couldn't reproduce it I think I was using very small test > images from libsdl2-gfx) > > I don't know how reliably reproducible this is. Any Flatpak app that uses > the documents portal and does not have direct filesystem access should be > equally suitable. Thanks for investing time into it and good at least you triggered the issue as well. > My laptop is not particularly suitable for building distro kernels, > but I'll try... please don't wait for me though! Don't bother likely if it's too much hassle to build the kernel. I have uploaed this morning 5.15.5-1 to unstable and cherry-picked the mentioned commit as well. But it's currently waiting for processing from NEW due to the ABI bump. https://salsa.debian.org/kernel-team/linux/-/commit/3f0763144c66a873774943ad335ddd34da8daed3 Regards, Salvatore
[toc] | [prev] | [next] | [standalone]
| From | Simon McVittie <smcv@debian.org> |
|---|---|
| Date | 2021-12-06 12:10 +0100 |
| Message-ID | <Drk2B-1nL-1@gated-at.bofh.it> |
| In reply to | #73805 |
On Fri, 26 Nov 2021 at 13:55:40 +0100, Salvatore Bonaccorso wrote:
> I have uploaed this morning 5.15.5-1 to unstable and cherry-picked the
> mentioned commit as well.
This appears to work correctly as far as I can tell, and is included in
5.16-rc3, 5.15.6, 5.10.83, 5.4.163 and 4.19.219 upstream.
smcv
[toc] | [prev] | [next] | [standalone]
| From | "Debian Bug Tracking System" <owner@bugs.debian.org> |
|---|---|
| Date | 2021-11-26 17:10 +0100 |
| Subject | Bug#1000504: marked as done (linux-image-5.15.0-1-amd64: FUSE regression in 5.15.3 affecting xdg-desktop-portal) |
| Message-ID | <DnLXt-4rn-25@gated-at.bofh.it> |
| In reply to | #73779 |
[Multipart message — attachments visible in raw view] — view raw
Your message dated Fri, 26 Nov 2021 16:00:11 +0000 with message-id <E1mqddz-0003oT-KL@fasolo.debian.org> and subject line Bug#1000504: fixed in linux 5.15.5-1 has caused the Debian Bug report #1000504, regarding linux-image-5.15.0-1-amd64: FUSE regression in 5.15.3 affecting xdg-desktop-portal to be marked as done. This means that you claim that the problem has been dealt with. If this is not the case it is now your responsibility to reopen the Bug report if necessary, and/or fix the problem forthwith. (NB: If you are a system administrator and have no idea what this message is talking about, this may indicate a serious mail system misconfiguration somewhere. Please contact owner@bugs.debian.org immediately.) -- 1000504: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1000504 Debian Bug Tracking System Contact owner@bugs.debian.org with problems
[toc] | [prev] | [standalone]
Back to top | Article view | linux.debian.kernel
csiph-web