Re: [PATCH] linux-user/elfload: use MAP_FIXED in pgb_reserved_va

[Alex Bennée]

Peter Maydell <peter.maydell@linaro.org> writes:

> On Tue, 30 Jun 2020 at 11:36, Alex Bennée <alex.bennee@linaro.org> wrote:
>>
>> Given we assert the requested address matches what we asked we should
>> also make that clear in the mmap flags. Otherwise we see failures in
>> the GitLab environment for some currently unknown but allowable
>> reason.
>
> Adding MAP_FIXED will mean that instead of failing if there's
> something else already at that address, the kernel will now
> silently blow that away in favour of the new mapping. Is
> that definitely what we want here ?

Hmm maybe not. But hey I just noticed that we have MAP_FIXED_NOREPLACE
(since Linux 4.17) which says:

   This flag provides behavior that is similar  to  MAP_FIXED  with
   respect   to   the   addr   enforcement,  but  differs  in  that
   MAP_FIXED_NOREPLACE never clobbers a preexisting  mapped  range.
   If  the  requested range would collide with an existing mapping,
   then this call fails with  the  error  EEXIST.   This  flag  can
   therefore  be used as a way to atomically (with respect to other
   threads) attempt to map an address range: one thread  will  suc‐
   ceed; all others will report failure.

   Note   that   older   kernels   which   do   not  recognize  the
   MAP_FIXED_NOREPLACE flag will typically (upon detecting a colli‐
   sion  with a preexisting mapping) fall back to a "non-MAP_FIXED"
   type of behavior: they will return an address that is  different
   from  the  requested  address.   Therefore,  backward-compatible
   software should check the returned address against the requested
   address.

So maybe that is what we should do?

Now you've pointed that out I wonder if we need to fix
pgd_find_hole_fallback as well?

>
> thanks
> -- PMM


-- 
Alex Bennée