[Mono-dev] How To Convince Mono To Allocate Big Arrays

[Rodrigo Kumpera]

Hi Luis,

Most parts of your patch are ok to be commited. As I said before, it's
better to have separate patches to speed up the review. As you get more
changes in, it will ease your work on preparing patches and will ease on
tracking head changes.

Right now I'm ok with the configure, mono_array_t and related changes. These
were easy to review and don't introduce any change in behavior, so we can
check-in then right now. The others requires more testing from me and would
be easier to both of us if done after.
Do you mind cooking you a patch only with the ok parts.

Remember that you need to either release these changes under the X11 license
or have done some copyright assignment paperwork.

Thanks,
Rodrigo



On Thu, May 8, 2008 at 8:21 PM, Luis F. Ortiz <LuisOrtiz at verizon.net> wrote:

> On May 8, 2008, at 9:29 AM, Rodrigo Kumpera wrote:
>
>> One important thing I forgot. If you break your patch into a few smaller
>> ones the review process will be a lot easier to every one involved.
>> The first one can introduce new types and configuration foo; then other to
>> fix codegen and Array methods and; at last, a bunch of fixes to classlib
>> issues -like sockets, file i/o and so on.
>>
>> Cheers,
>> Rodrigo
>>
>
> OK, here is another stab at the changes.
>
> This modifies the following files:
>        configure.in
>        mono/metadata/object.c
>        mono/metadata/object.h
>        mono/metadata/icall-def.h
>        mono/metadata/icall.c
>        mono/metadata/socket-io.c
>
> These files:
>  1) Add a new configuration option, --enable-big-arrays, which defines
> conditionally defines MONO_BIG_ARRAYS in config.h,
>  2) Add in mono/metadata/object.h :
>  A) A new typedef for mono_array_size_t to be either guint32 or guint64
>  B) A new #define for MONO_ARRAY_MAX_INDEX for the biggest valid array
> index, i.e. G_MAXINTxx
>  C) A new #define for MONO_ARRAY_MAX_SIZE for the biggest valid array
> allocation, i.e. G_MAXUINTxx
>  D) The above all controlled by MONO_BIG_ARRAYS
>  3) Modify the definitions of mono_array_new(), mono_array_new_full(), and
> mono_array_new_specific() to match,
>  4) Modify the usages of those functions in the metadata directory to
> match,
>  5) Add range checking in ves_icall_System_Array_CreateInstanceImpl64 so it
> works with or without MONO_BIG_ARRAYS,
>  6) Attempt to address all the mistakes you pointed out.
>
> These changes, in addition to the other changes I made, pass "make check"
> with the exception of Tests.test_0_regress_75832(), as previously confessed.
>
>
>
>
>
> /Ortiz/Luis
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ximian.com/pipermail/mono-devel-list/attachments/20080516/948d959e/attachment-0001.html