ffc team mailing list archive

[Andy Ray Terrel <andy.terrel@xxxxxxxxx>]

On Mon, Aug 8, 2011 at 11:46 AM, Anders Logg <logg@xxxxxxxxx> wrote:
> On Mon, Aug 08, 2011 at 10:38:37AM -0500, Andy Ray Terrel wrote:
>> On Mon, Aug 8, 2011 at 10:27 AM, Johan Hake <johan.hake@xxxxxxxxx> wrote:
>> > On Monday August 8 2011 04:33:41 Anders Logg wrote:
>> >> On Wed, Aug 03, 2011 at 08:49:24PM -0700, Johan Hake wrote:
>> >> > Hello!
>> >> >
>> >> > Why do we write generated code to file when the jit compilation is used?
>> >> > This introduce a cause of error which has annoyed me for some time now.
>> >> > Instead of writing the code to file and then in ufc.build we just read
>> >> > the file again, we could just pass the code as a single str.
>> >> >
>> >> > This would save us for a lot of hassle with file access during server
>> >> > run. When several runs are commited and executed at one time I
>> >> > experience a lot of runs that stops prematurely because of file access
>> >> > problem. This "should" have been fixed by some file locking mechanism
>> >> > which really is not working.
>> >> >
>> >> > If you think this is a cool thing to do I can do the work. We just have
>> >> > to add a parameter, or an extra argument to some ffc functions and
>> >> > change the interface to ufc.build.
>> >>
>> >> One advantage is that it's possible to inspect the generated code in
>> >> ~/.instant/cache if it fails.
>> >
>> > That is true. But this can be accomplished by looking into the generated .i
>> > file instead.
>>
>> As FEniCS tries to run on bigger machines, the need to not write to
>> files becomes more important.  Many of the top machines do not have
>> disk and other things.  I think this feature would be necessary for
>> those machines.
>
> But isn't the problem still there? We still need to write the .i file,
> and gcc spits out a .so file. And other files are still being written
> to the cache directory, so I don't see the big benefit in removing
> only one of those files.

Yup those files would have to be addressed as well.  Just throwing my
2 cents in because I have been bit by all the files FEniCS writes out.

-- Andy

>
> --
> Anders
>
>
>> -- Andy
>>
>> >
>> > Johan
>> >
>> > _______________________________________________
>> > Mailing list: https://launchpad.net/~ffc
>> > Post to     : ffc@xxxxxxxxxxxxxxxxxxx
>> > Unsubscribe : https://launchpad.net/~ffc
>> > More help   : https://help.launchpad.net/ListHelp
>> >
>>
>> _______________________________________________
>> Mailing list: https://launchpad.net/~ffc
>> Post to     : ffc@xxxxxxxxxxxxxxxxxxx
>> Unsubscribe : https://launchpad.net/~ffc
>> More help   : https://help.launchpad.net/ListHelp
>