On 24 July 2018 at 17:29, Sebastian Andrzej Siewior
<bigeasy@xxxxxxxxxxxxx> wrote:
> On 2018-07-24 17:00:09 [+0200], Peter Zijlstra wrote:
>> On Tue, Jul 24, 2018 at 04:35:09PM +0200, Sebastian Andrzej Siewior wrote:
>> > I doubt that there any need to set ->active_mm. It is used by the
>> > scheduler to keep track of the "currently used mm" so it can reuse one
>> > for the kernel thread which does not own one and take a reference on it
>> > so it does not go away while the thread (that borrows it) is active.
>>
>> > void efi_switch_mm(struct mm_struct *mm)
>> > {
>> > - task_lock(current);
>> > efi_scratch.prev_mm = current->active_mm;
>> > - current->active_mm = mm;
>> > switch_mm(efi_scratch.prev_mm, mm, NULL);
>> > - task_unlock(current);
>> > }
>>
>> I think that's broken. Take for instance stuff like
>> perf_callchain_user32() -> get_segment_base(). That looks at active_mm
>> to get at the current LDT.
>
> right. I saw that briefly not sure why I dropped it. I have no idea
> where the LDT points to but it probably sense to return EFI's version of
> it.
>
>> Now, I'm not saying the whole perf vs EFI thing isn't already terminally
>> wrecked, but the rule is that active_mm really should point at the
>> current active mm, and the above breaks that.
> Right. Even if we not perform a context switch. Okay. Will update that
> part.
>
Please refer to what has been queued up in tip:efi/core. Sai has
implemented a work queue for EFI calls so they occur from a kernel
thread, and the mixed mode locking has been fixed as well.
--
To unsubscribe from this list: send the line "unsubscribe linux-efi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
