> That would be very interesting.
>
> Danny
>
I forgot to say that there is something more important that releasing a new
cegcc version
it's to find why it doesn't work on recent device under WM 6.1.
We are working with Jerome Decoodt to solve this issue but it's not very
easy because we nee
On Sat, 2008-08-30 at 11:43 +0200, Vincent Torri wrote:
> >> Both these trees are inherited, it might be worth looking at how they
> >> evolved since our "fork".
> >
> > Indeed, no d_type too in the mingw header. I'll ask in the mingw ML, then
>
> btw, a thread in the mingw ML that was started 2 m
On Sat, 2008-08-30 at 12:26 +0200, [EMAIL PROTECTED] wrote:
> Last time I tried to report modifications on a GCC 4.3.1 I got a compiler
> error when compiling libstdc++ but I could always tried
> it with this new release and keep you informed.
That would be very interesting.
Danny
--
Dan
Great!
It's hard to say if GCC 4.3.2 is a good version for cegcc because as you
may now a developper(Aaron W. LaFramboise) is currently working hard
to improve GCC for windows and especially in the area of exceptions and
"DLLification".
Now if you look at patch submitted to GCC :
Re: [PATCH]
On Sat, 30 Aug 2008, Vincent Torri wrote:
>
>
> On Sat, 30 Aug 2008, Danny Backx wrote:
>
>> On Sat, 2008-08-30 at 11:08 +0200, Vincent Torri wrote:
>>> Hey,
>>>
>>> on unix, the dirent structure contains the memberd_type, that store the
>>> type of file (dir, link, etc...), and macro values (th
On Sat, 30 Aug 2008, Danny Backx wrote:
> On Sat, 2008-08-30 at 11:08 +0200, Vincent Torri wrote:
>> Hey,
>>
>> on unix, the dirent structure contains the memberd_type, that store the
>> type of file (dir, link, etc...), and macro values (the DT_*)
>>
>> They do not exist in the cegcc dirent.h f
On Sat, 30 Aug 2008, Danny Backx wrote:
> Pedro has added a lot of stuff to our mingw library. After I first
> joined the project, my impression was that he evolved from supporting
> both cegcc and ming32ce to focussing on the latter.
>
> I must say I gradually moved in that direction too.
>
> S
On Sat, 2008-08-30 at 11:08 +0200, Vincent Torri wrote:
> Hey,
>
> on unix, the dirent structure contains the memberd_type, that store the
> type of file (dir, link, etc...), and macro values (the DT_*)
>
> They do not exist in the cegcc dirent.h file. Is it normal ?
As I said in the other mess
On Sat, 2008-08-30 at 10:21 +0200, Vincent Torri wrote:
> > The most annoying to me is I couldn't get it (4.3.0) to work.
>
> the c++ stuff ?
Yes.
> >> also a discussion about a new newlib would be nice
> >
> > How has it evolved since the version we currently have ?
>
> cegcc rely on newlib 1.
Hey,
on unix, the dirent structure contains the memberd_type, that store the
type of file (dir, link, etc...), and macro values (the DT_*)
They do not exist in the cegcc dirent.h file. Is it normal ?
Vincent Torri
-
This
On Sat, 30 Aug 2008, Danny Backx wrote:
> On Sat, 2008-08-30 at 01:26 +0200, Vincent Torri wrote:
>>
>> On Fri, 29 Aug 2008, Danny Backx wrote:
>>
>>> Maybe this can re-open the discussion about gcc versions in cegcc.
>>>
>>> Which version should be next ?
>>
>> gcc 4.3.3 will be out end of octo
On Sat, 2008-08-30 at 01:26 +0200, Vincent Torri wrote:
>
> On Fri, 29 Aug 2008, Danny Backx wrote:
>
> > Maybe this can re-open the discussion about gcc versions in cegcc.
> >
> > Which version should be next ?
>
> gcc 4.3.3 will be out end of october. Is there some annoying bugs in 4.3.2
> ?
This required an additional parameter to FDCHECK.
Committed too.
Danny
On Fri, 2008-08-29 at 15:52 -0700, Pawel Veselov wrote:
> Hi,
>
> this attached fix should take care of providing fsync()/fdatasync()
> functionliaty. For now it's a straight call to FlushFileBuffers().
>
> Thanks,
>
On Fri, 2008-08-29 at 22:53 -0700, Pawel Veselov wrote:
> The entry should said something like:
> - prevent abandoning critical section lock when calling close() on an
> invalid file descriptor.
I've committed.
> Let me know if you rather I submit the change log as part of the
> patch.
For futur
14 matches
Mail list logo
