Re: [PATCH 1/8] ucounts: Fix RLIMIT_NPROC regression

[Eric W. Biederman]

Michal Koutný <mkoutny@xxxxxxxx> writes:

> On Thu, Feb 10, 2022 at 08:13:17PM -0600, "Eric W. Biederman" <ebiederm@xxxxxxxxxxxx> wrote:
>> This can be fixed either by fixing the test or by moving the increment
>> to be before the test.  Fix it my moving copy_creds which contains
>> the increment before is_ucounts_overlimit.
>
> This is simpler than my approach and I find it correct too.
>
>> Both the test in fork and the test in set_user were semantically
>> changed when the code moved to ucounts.  The change of the test in
>> fork was bad because it was before the increment.
>>
>> The test in set_user was wrong and the change to ucounts fixed it.  So
>> this fix is only restore the old behavior in one lcatio not two.
>
> Whom should be the task accounted to in the case of set*uid? (The change
> to ucounts made the check against the pre-switch user's ucounts.)

It needs to be post-switch in the case of set*id().

I have that fixed in the next version of my patchset.

>> ---
>>  kernel/fork.c | 10 +++++-----
>>  1 file changed, 5 insertions(+), 5 deletions(-)
>
> Reviewed-by: Michal Koutný <mkoutny@xxxxxxxx>