[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#1074170: marked as done (linux: missing 5.10 backport of pfn_valid() changes from f6564fce256a3944aa1bc76cb3c40e792d97c1eb ?)

Your message dated Sat, 22 Feb 2025 21:17:07 +0100
with message-id <Z7oww9Sx1UTaCv8j@eldamar.lan>
and subject line Re: Bug#1074170: linux: missing 5.10 backport of pfn_valid() changes from f6564fce256a3944aa1bc76cb3c40e792d97c1eb ?
has caused the Debian Bug report #1074170,
regarding linux: missing 5.10 backport of pfn_valid() changes from f6564fce256a3944aa1bc76cb3c40e792d97c1eb ?
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.)


-- 
1074170: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1074170
Debian Bug Tracking System
Contact owner@bugs.debian.org with problems
--- Begin Message --- 
Source: linux
Version: 5.10.218-1
Severity: important
Tags: upstream

There are two commits in 6.8 that modify (the arch independent)
pfn_valid() in include/linux/mmzone.h to fix race conditions:

mm/sparsemem: fix race in accessing memory_section->usage
5ec8e8ea8b7783fab150cf86404fc38cb4db8800 (v6.8-rc1)
introduces usage of rcu_read_lock()/rcu_read_unlock()
(which are (transitively) GPL-only symbols)

mm, kmsan: fix infinite recursion due to RCU critical section
f6564fce256a3944aa1bc76cb3c40e792d97c1eb (v6.8-rc3)
switches that to rcu_read_lock_sched()/rcu_read_unlock_sched()
(which are not)

Both commits got backported to Linux 6.1 (in bookworm) in
v6.1.76/v6.1.77 but so far only the first got backported to Linux 5.10
(in bullseye) in v5.10.210.

This triggered a module build failure of the proprietary nvidia driver
in bullseye-pu on ppc64el (no other architectures affected), while it
succeeds for the kernels in bookworm-pu and sid.
(see #1072965 for details)


Andreas

--- End Message ---
--- Begin Message --- 
Hi,

On Mon, Jun 24, 2024 at 10:58:12AM +0200, Andreas Beckmann wrote:
> Source: linux
> Version: 5.10.218-1
> Severity: important
> Tags: upstream
> 
> There are two commits in 6.8 that modify (the arch independent)
> pfn_valid() in include/linux/mmzone.h to fix race conditions:
> 
> mm/sparsemem: fix race in accessing memory_section->usage
> 5ec8e8ea8b7783fab150cf86404fc38cb4db8800 (v6.8-rc1)
> introduces usage of rcu_read_lock()/rcu_read_unlock()
> (which are (transitively) GPL-only symbols)
> 
> mm, kmsan: fix infinite recursion due to RCU critical section
> f6564fce256a3944aa1bc76cb3c40e792d97c1eb (v6.8-rc3)
> switches that to rcu_read_lock_sched()/rcu_read_unlock_sched()
> (which are not)
> 
> Both commits got backported to Linux 6.1 (in bookworm) in
> v6.1.76/v6.1.77 but so far only the first got backported to Linux 5.10
> (in bullseye) in v5.10.210.
> 
> This triggered a module build failure of the proprietary nvidia driver
> in bullseye-pu on ppc64el (no other architectures affected), while it
> succeeds for the kernels in bookworm-pu and sid.
> (see #1072965 for details)

Given bullseye is in LTS now, and does not support ppc64el I think we
can close this bug now.

Regards,
Salvatore

--- End Message ---
Reply to: