Re: CVE-2022-48921: sched/fair: Fix fault in reweight_entity

Re: CVE-2022-48921: sched/fair: Fix fault in reweight_entity

[Xia Fukun]

On 2024/8/22 11:31, Greg Kroah-Hartman wrote:
> Description
> ===========
> 
> In the Linux kernel, the following vulnerability has been resolved:
> 
> sched/fair: Fix fault in reweight_entity
> 
> Syzbot found a GPF in reweight_entity. This has been bisected to
> commit 4ef0c5c6b5ba ("kernel/sched: Fix sched_fork() access an invalid
> sched_task_group")
> 
> There is a race between sched_post_fork() and setpriority(PRIO_PGRP)
> within a thread group that causes a null-ptr-deref in
> reweight_entity() in CFS. The scenario is that the main process spawns
> number of new threads, which then call setpriority(PRIO_PGRP, 0, -20),
> wait, and exit.  For each of the new threads the copy_process() gets
> invoked, which adds the new task_struct and calls sched_post_fork()
> for it.
> 
> 
> The Linux kernel CVE team has assigned CVE-2022-48921 to this issue.
> 

Commit 13765de8148f ("sched/fair: Fix fault in reweight_entity")
is reverted by commit b1e8206582f9 ("sched: Fix yet more sched_fork()
races"). Since commit 13765de8148f only fixes a single instance
of this problem, not the whole class.

I think the CVE-2022-48921 needs to adjust the corresponding commit
to commit b1e8206582f9 ("sched: Fix yet more sched_fork() races").

Re: CVE-2022-48921: sched/fair: Fix fault in reweight_entity

[Greg Kroah-Hartman]

On Sat, Aug 24, 2024 at 05:52:05PM +0800, Xia Fukun wrote:
> 
> On 2024/8/22 11:31, Greg Kroah-Hartman wrote:
> > Description
> > ===========
> > 
> > In the Linux kernel, the following vulnerability has been resolved:
> > 
> > sched/fair: Fix fault in reweight_entity
> > 
> > Syzbot found a GPF in reweight_entity. This has been bisected to
> > commit 4ef0c5c6b5ba ("kernel/sched: Fix sched_fork() access an invalid
> > sched_task_group")
> > 
> > There is a race between sched_post_fork() and setpriority(PRIO_PGRP)
> > within a thread group that causes a null-ptr-deref in
> > reweight_entity() in CFS. The scenario is that the main process spawns
> > number of new threads, which then call setpriority(PRIO_PGRP, 0, -20),
> > wait, and exit.  For each of the new threads the copy_process() gets
> > invoked, which adds the new task_struct and calls sched_post_fork()
> > for it.
> > 
> > 
> > The Linux kernel CVE team has assigned CVE-2022-48921 to this issue.
> > 
> 
> Commit 13765de8148f ("sched/fair: Fix fault in reweight_entity")
> is reverted by commit b1e8206582f9 ("sched: Fix yet more sched_fork()
> races"). Since commit 13765de8148f only fixes a single instance
> of this problem, not the whole class.
> 
> I think the CVE-2022-48921 needs to adjust the corresponding commit
> to commit b1e8206582f9 ("sched: Fix yet more sched_fork() races").

I think we just need to assign a new CVE to b1e8206582f9, as that was
not backported to everywhere that 13765de8148f was applied, right?
Wouldn't that be the correct thing to do as it did fix things in a
different way.

thanks,

greg k-h

Re: CVE-2022-48921: sched/fair: Fix fault in reweight_entity

[Greg Kroah-Hartman]

On Sun, Aug 25, 2024 at 07:54:40AM +0200, Greg Kroah-Hartman wrote:
> On Sat, Aug 24, 2024 at 05:52:05PM +0800, Xia Fukun wrote:
> > 
> > On 2024/8/22 11:31, Greg Kroah-Hartman wrote:
> > > Description
> > > ===========
> > > 
> > > In the Linux kernel, the following vulnerability has been resolved:
> > > 
> > > sched/fair: Fix fault in reweight_entity
> > > 
> > > Syzbot found a GPF in reweight_entity. This has been bisected to
> > > commit 4ef0c5c6b5ba ("kernel/sched: Fix sched_fork() access an invalid
> > > sched_task_group")
> > > 
> > > There is a race between sched_post_fork() and setpriority(PRIO_PGRP)
> > > within a thread group that causes a null-ptr-deref in
> > > reweight_entity() in CFS. The scenario is that the main process spawns
> > > number of new threads, which then call setpriority(PRIO_PGRP, 0, -20),
> > > wait, and exit.  For each of the new threads the copy_process() gets
> > > invoked, which adds the new task_struct and calls sched_post_fork()
> > > for it.
> > > 
> > > 
> > > The Linux kernel CVE team has assigned CVE-2022-48921 to this issue.
> > > 
> > 
> > Commit 13765de8148f ("sched/fair: Fix fault in reweight_entity")
> > is reverted by commit b1e8206582f9 ("sched: Fix yet more sched_fork()
> > races"). Since commit 13765de8148f only fixes a single instance
> > of this problem, not the whole class.
> > 
> > I think the CVE-2022-48921 needs to adjust the corresponding commit
> > to commit b1e8206582f9 ("sched: Fix yet more sched_fork() races").
> 
> I think we just need to assign a new CVE to b1e8206582f9, as that was
> not backported to everywhere that 13765de8148f was applied, right?
> Wouldn't that be the correct thing to do as it did fix things in a
> different way.

CVE-2022-48944 is now assigned for this, thanks.

greg k-h