[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 25 Aug 2015 17:11:04 +0200
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: 7z-opencl crashes with auto-tuning
On 2015-08-25 14:39, Solar Designer wrote:
> On Tue, Aug 25, 2015 at 03:27:47PM +0300, Solar Designer wrote:
>> The 7z-opencl format passes test when run with explicit GWS and LWS, but
>> crashes during/after my patched auto-tuning:
>
> When targeting -dev=1, rar-opencl similarly crashes with the auto-tuning
> patch, at the same point (when starting to tune GWS for the second
> time). However, surprisingly it doesn't crash on -dev=5 (and actually
> gets tuned better than before). 7z-opencl crashes on both.
>
> There are probably changes needed to these two formats, or to how the
> auto-tuning is revised - or both.
>
> Maybe more formats are broken with that patch. I only tested a few.
This (committed) fix should fix all formats with similar problem.
diff --git a/src/common-opencl.c b/src/common-opencl.c
index 7f0440c..b86a7ce 100644
--- a/src/common-opencl.c
+++ b/src/common-opencl.c
@@ -1313,6 +1313,9 @@ static cl_ulong gws_test(size_t gws, unsigned int
rounds, int sequential_id)
for (i = 0; i < MAX_EVENTS; i++)
benchEvent[i] = NULL;
+ // Ensure format knows its GWS
+ global_work_size = gws;
+
// Prepare buffers.
create_clobj(gws, self);
The canonical fix would be a major clean-up of which variables should be
global (and perhaps prefixed with ocl_) and which should be format-local
(like global_work_size). It's a mess, for historical reasons.
magnum
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
