Re: [PATCH 00/13] Re: Scalability requirements for sysv ipc

From: Nadia Derbey
Date: Mon Apr 14 2008 - 06:52:40 EST


Nadia Derbey wrote:
Peter Zijlstra wrote:

On Mon, 2008-04-14 at 07:18 +0200, Nadia Derbey wrote:

Peter Zijlstra wrote:

On Fri, 2008-04-11 at 18:17 +0200, Nadia.Derbey@xxxxxxxx wrote:


Here is finally the ipc ridr-based implementation I was talking about last
week (see http://lkml.org/lkml/2008/4/4/208).
I couldn't avoid much of the code duplication, but at least made things
incremental.

Does somebody now a test suite that exists for the idr API, that I could
run on this new api?

Mike, can you try to run it on your victim: I had such a hard time building
this patch, that I couldn't re-run the test on my 8-core with this new
version. So the last results I have are for 2.6.25-rc3-mm1.

Also, I think a careful review should be done to avoid introducing yet other
problems :-(



Why duplicate the whole thing, when we converted the Radix tree to be
RCU safe we did it in-place. Is there a reason this is not done for idr?




I did that because I wanted to go fast and try to fix the performance problem we have with sysV ipc's. I didn't want to introduce (yet other) regressions in the code that uses idr's today and that works well ;-)
May be in the future if this rcu based api appears to be ok, we can replace one with the other?



From what I can see the API doesn't change at all,


Well, 1 interface changes, 1 is added and another one went away:

1) for the preload part (it becomes like the radix-tree preload part):

int idr_pre_get(struct idr *, gfp_t);
would become
int idr_pre_get(gfp_t);

2) idr_pre_get_end() is added (same as radix_tree_preload_end()).

3) The idr_init() disappears.

You might see that other interfaces are not provided by ridr, but this is only because I've taken those that are useful for the ipc part (so should not be a problem to make the whole thing rcu safe).

so I don't see why
you need to duplicate - either the new code works as expected or its
broken.


That's why I asked for an "IDR test suite": I wanted to test potential regressions.

If it works its good enough for all IDR users, if its broken we
should fix it. Seems simple enough.. am I missing something obvious?


Regards,
Nadia


BTW, I'm realizing that I forgot to send the results I've got - sorry for that (I finally could pass the pmsg/psem tests this morning):

These are the output files for the command:

for i in 1 2 3 4 5 6 7 8;do ./pmsg $i 5;done

pmsg_output.25_rc8_mm1.ref.8: output file for the 2.6.25-rc8-mm1
reference kernel
pmsg_output.25_rc8_mm1.ridr.8: output file for the 2.6.25-rc8-mm1
refrence kernel, with ipc's using
rcu-based idr's
psem_output.25_rc8_mm1.ref.8: same as <1> for the psem test
psem_output.25_rc8_mm1.ridr.8: same as <2> for the psem test

Regards,
Nadia

Attachment: results.tar.bz2
Description: application/redhat-package-manager