Hi guys,
'putc' in mingw/include/stdio.h was missing a 'return'.
Fixed in trunk with the patch below.
Cheers,
Pedro Alves
---
2006-10-22 Pedro Alves <[EMAIL PROTECTED]>
* include/stdio.h (puts) [__COREDLL__]: Add missing return
alk, patch below.
Cheers,
Pedro Alves
---
src/mingw/ChangeLog.mingw32ce
2006-10-22 Pedro Alves <[EMAIL PROTECTED]>
* configure.in [AC_CONFIG_SUBDIRS]: Add mingwex.
* configure : Regenerate.
* mingwex/Makefile.in (VPATH): Add wince subdir.
(WINCE_DISTFILES): New.
(
Hi all,
My latest mingwex commit enables a bit of cleaning up of libstdc++.
Also, since these functions were currently disabled,
building a libstdc++-6.dll with src/script.sh was broken, as full
closure was
not possible.
Cheers,
Pedro Alves
---
src/gcc/libstdc++-v3/ChangeLog.ce
2006-10-22
Hi all,
I just committed a patch that enables test_headers.c to compile with
current mingw32ce. This is useful for mingw/include header testing.
Cheers,
Pedro Alves
---
2006-10-22 Pedro Alves <[EMAIL PROTECTED]>
* test_headers.c: Adapt to __COREDLL__.
Index: test_hea
t for it, but it doesn't mention anything about
the ABI being incompatible.
Cheers,
Pedro Alves
-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrat
Hi Andy,
Andy Wu escreveu:
> Hi Pedro,
>
>
> On 10/24/06, *Pedro Alves* < [EMAIL PROTECTED]
> <mailto:[EMAIL PROTECTED]>> wrote:
>
> Hi Andy,
>
> Andy Wu wrote:
>
> > Hi all,
> >
> > The w
arm emulator from MS for free:
http://www.microsoft.com/downloads/details.aspx?familyid=C62D54A5-183A-4A1E-A7E2-CC500ED1F19A&displaylang=en
There is an page on the old berlios Wiki about this:
http://cegcc.berlios.de/wi
Trevor MacPhail wrote:
> On 25-Oct-06, at 11:28 AM, Pedro Alves wrote:
>
>
>> But, why bother? You can get an arm emulator from MS for free:
>> http://www.microsoft.com/downloads/details.aspx?
>> familyid=C62D54A5-183A-4A1E-A7E2-CC500ED1F19A&displaylang=
Trevor MacPhail wrote:
> On 25-Oct-06, at 11:28 AM, Pedro Alves wrote:
>
>
>> But, why bother? You can get an arm emulator from MS for free:
>> http://www.microsoft.com/downloads/details.aspx?
>> familyid=C62D54A5-183A-4A1E-A7E2-CC500ED1F19A&displaylang=
Pedro Alves wrote:
> The float ABI is something that has been on my todo list,
> like confirming the alignments, packing, endianess, calling conventions
> related to floats/doubles.
> I would really like to find some authorative (MSDN/Microsoft)
> information about this.
>
g Windows look like Linux/Unix.
You could also try the src/build-cegcc.sh for cegcc, and
src/build-mingw32ce.sh for mingw32ce.
I regularly use them to build under linux and cygwin without change.
Cheers,
Pedro Alves
-
Using Tomcat but
Pedro Alves wrote:
Pedro Alves wrote:
The float ABI is something that has been on my todo list,
like confirming the alignments, packing, endianess, calling conventions
related to floats/doubles.
I would really like to find some authorative (MSDN/Microsoft)
information about this
e let me know that.
>
>
>
This was just answered yesterday, but since the mailing list archives
have been somewhat flacky, I'll
cut you some slack ;)
For cegcc:
src/build-cegcc.sh
For mingw32ce:
src/build-mingw32ce.sh
>thank you for reading a
into src/ and build from there using those scripts.
This will get you a working toolchain much faster than waiting for the
next cygwin binary release.
Cheers,
Pedro Alves
>again thank you for your reply and im gonna wait for your answer.
>
the exact steps for building the
full cegcc toolchain.
Note that if you don't want the newlib dependency, you really want
mingw32ce, not cegcc.
Cheers,
Pedro Alves
-
Using Tomcat but need to do more? Need to support w
ould be
pretty easy to at least get the app to build,
and minimally run.
Then, evaluate if you really need newlib/cegcc.dll. Porting to mingw32ce
is a bit harder, as more stuff is missing.
Another good approach, would be to port the app first to desktop MinGW,
and then port it to mingw32ce.
Of cou
Hi all,
I just committed this patch to the build scripts to fix the bug in
build-mingw32ce.sh that made the build continue for Carl even though
an error had occurred.
Also, I added a few missing error handlers in build-cegcc.sh and
made a bit of house cleaning.
Cheers,
Pedro Alves
---
src
Pedro Alves wrote:
> Hi all,
>
> I just committed this patch to the build scripts to fix the bug in
> build-mingw32ce.sh that made the build continue for Carl even though
> an error had occurred.
Obviously, to build-cegcc.sh, not build-mingw32ce.sh.
Chee
gt; Also, I need to build the c++ compiler. I noticed the default scripts
> don't build this for the mingw environment. Is this going to be
> difficult to get running?
>
You can also try building using the alternative scripts under src/
c
l support is doable, but it requires hacking the gdb's remote
protocol. Since desktop MinGW and cygwin would benefit from
this work too, we might get some good help from the gdb folks.
Cheers,
Pedro Alves
-
Using Tomcat but
Pedro Alves wrote:
>Not much time at the moment to analise this, quite swamped at work.
>
>Nuno Lucas wrote:
>
>
>
>>On 10/29/06, Kevin O'Connor <[EMAIL PROTECTED]> wrote:
>>
>>
>>
>>>Okay. That seems to work. The build still
bin/arm-wince-mingw32ce-as "$@"
Maybe it is being called for some reason?
Maybe you could try adding a '-v' to CFLAGS, so we can understand why
the wrong as is being called?
Cheers,
Pedro Alves
-
Using T
Danny Backx wrote:
>On Mon, 2006-10-30 at 12:23 +0000, Pedro Alves wrote:
>
>
>>>The real solution would be to port gdbserver (there is a mingw port in
>>>gdb-cvs) to wince. The problem is that
>>>gdbserver doesn't have dll debugging support. If that
rsion number to it so we can have multiple versions. Also, with these numbers
attached,
there is no conflict with current cegcc's libsup++ and libstdc++ dlls, so they
can be installed
side by side on the device.
Cheers,
Pedro Alves
/
--
at should be
submitted upstream.
I'm only posting the ChangeLog, since the patch is pretty big. You can
see it at cegcc-svn
or by svn updating.
I only tested this on mingw32ce yet, will test cegcc tomorrow.
Cheers,
Pedro Alves
---
src/gcc/gcc/ChangeLog.ce
2006-10-30 Pedro Alves <[EM
ecorded as armv5.
Cheers,
Pedro Alves
---
src/binutils/gas/ChangeLog.ce
2006-10-31 Pedro Alves <[EMAIL PROTECTED]>
* config/te-wince-pe-h (CPU_DEFAULT): Define to ARM_ARCH_V4.
(FPU_DEFAULT) Default to FPU_ARCH_VFP.
Index: gas/conf
Pedro Alves
---
bfd/ChangeLog
2006-10-21 Pedro Alves <[EMAIL PROTECTED]>
* pe-arm-wince.c (LOCAL_LABEL_PREFIX): Define as ".".
* pei-arm-wince.c (LOCAL_LABEL_PREFIX): Likewise.
* coff-arm.c (LOCAL_LABEL_PREFIX): Only define if not defined before.
gas/ChangeLog
2006-10-2
Alves
ld/ChangeLog
2006-10-29 Pedro Alves <[EMAIL PROTECTED]>
* pe-dll.c (make_singleton_name_thunk): Re-add the NULL terminator.
-
Using Tomcat but need to do more? Need to support web services, security?
Get
I'm puzzled.
I just built cegcc and mingw32ce to test for the recent patches,
and both versions of gdb built perfectly.
I used the scripts under src/
Cheers,
Pedro Alves
-
Using Tomcat but need to do more? Need to su
Nuno Lucas wrote:
>On 10/31/06, Pedro Alves <[EMAIL PROTECTED]> wrote:
>
>
>>I'm puzzled.
>>I just built cegcc and mingw32ce to test for the recent patches,
>>and both versions of gdb built perfectly.
>>I used the scripts under src/
>>
>
Pedro Alves wrote:
> Nuno Lucas wrote:
>
>> On 10/31/06, Pedro Alves <[EMAIL PROTECTED]> wrote:
>>
>>
>>> I'm puzzled.
>>> I just built cegcc and mingw32ce to test for the recent patches,
>>> and both versions of gdb built perfectly
Kevin O'Connor wrote:
> On Tue, Oct 31, 2006 at 07:03:45PM +0000, Pedro Alves wrote:
>
>> I'm puzzled.
>> I just built cegcc and mingw32ce to test for the recent patches,
>> and both versions of gdb built perfectly.
>> I used the scripts under src/
>
2api/include, we use _WIN32_WCE not UNDER_CE to check for WinCE.
UNDER_CE is defined, but not to any meaningful value.
I therefore committed the attached patch.
Thansks,
Cheers,
Pedro Alves
---
src/w32api/ChangeLog.ce:
2006-10-29 Kevin O'Connor <[EMAIL PROTECTED]>
* include/winuser.h
try this patch? I don't have time for anything else at the moment.
Cheers,
Pedro Alves
Index: pe.h
===
--- pe.h(revisão 769)
+++ pe.h(cópia de trabalho)
@@ -185,6 +185,7 @@
#define ASM_OUTPUT_EXTERNAL_LIBCALL(FILE, F
Simon Posnjak escreveu:
> Hi all
>
> I managed to build the svn version of the tool chains on Linux. I can
> confirm (as reported earlier) that there is a problem with building
> gdb on Ubuntu 6.10 - I attached the error.
>
I wonder where those non ascii characters ... :
/home/simonp/delo/cegcc/t
Hi all,
I'm committing the attached patch, to fix some listview related stuff,
as described here:
http://msdn2.microsoft.com/en-us/library/aa453423.aspx
Cheers,
Pedro Alves
2006-11-16 Pedro Alves <[EMAIL PROTECTED]>
* include/commctrl.h (LVCF_IMAGE, LVCF_ORDER) : Also
touches src/build-mingw32ce.sh, specifically, it changes
-I$(w32api) to -idirafter $(w32api),
to be sure win32 headers are always found after the system headers.
if you just want to build gdb, you can type:
'./build-mingw32ce.sh gdb'
followed by:
'./build-mingw32ce.sh gdbstub&#
vice was bad. Naively, I gave Pedro Admin rights after having setup
> the new site. And as of a couple of days ago, he abused those rights to
> take away my (1) mailing list admin, (2) project admin, and (3)
> developer access. If that isn't betrayal then I don't know what is
ader with ' extern "C" ' ... :
#ifndef _GX_H
#define _GX_H
#ifdef __cplusplus
extern "C"
{
#endif
#ifdef __cplusplus
}
#endif
#endif
(Take a look at the other headers in src/w32api/include in the sources,
or in /opt/cegcc/arm-wince-cegcc/include in the binary releases
worth it to be as compatible as possible with the
native tools.
Cheers,
Pedro Alves
gcc/gcc/ChangeLog.ce:
2006-11-08 Pedro Alves <[EMAIL PROTECTED]>
* config/arm/wince-pe.h (DEFAULT_SIGNED_CHAR): Define to 1.
Index: gcc/config/arm/winc
Danny Backx wrote:
>On Tue, 2006-11-07 at 20:41 +0000, Pedro Alves wrote:
>
>
>>You've got to be kidding... Your project?
>>
>>
>
>The sentence was meant to draw your attention to the fact that you're
>dealing with people here, not with slaves.
* mallocs 1 byte. */
}
In a mingw32ce app using the above dll:
__declspec(dllimport) struct small* alloc_small(void);
int main()
{
struct small* sm = alloc_small();
memset (sm, 0, sizeof (struct small)); /* clears 4 bytes, outch! */
}
> Op di 31-10-2006, om 00:59 schreef Pedro Alves:
>
Pedro Alves wrote:
> You can get a lot more info on this by googling for "-ms-bitfields" option.
>
For the record, you disable this globally with -mno-ms-bitfields,
or on a struct by struct basis using __attribute__((gcc_struct)), or
__attribute__((ms_struct)).
Che
, but,
I'm happy already,
Cheers,
Pedro Alves
-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebS
We've successfully sorted things out, and Danny is now back as a project
administrator,
with full privileges as before.
I'm very sorry that everyone had to see such an ugly discussion.
Hope you can all accept my apologies.
Cheers,
P
Hi all,
I've broken the building of src/mingw with revision 784
I have the fix here ready to commit, but my local trunk copy decided to
get broken.
I hope to be able to commit the fix in the next hours.
You can svn up to revision 783 to work around this.
Cheers,
Pedro
no(stdout), "msg", 3);
(*) -lmingwex is added by default by the compiler driver, you don't need
to do anything.
Cheers,
Pedro Alves
P.S.: I missed some things in the first patch, which I fixed on the second.
(patch 1)
src/mingw/ChangeLog.mingw32ce:
2006-11-11 Pedro Alves <[EMAIL
Pedro Alves wrote:
Hi all,
While porting gdbserver I stumbled on yet again having to provide
write and read functions
as wrappers around ReadFile and WriteFile.
This is about the fourth time I do in a short period, so, I've put
those functions
in libmingwex.a (*), to make this the last
Pedro Alves wrote:
> Hi all,
>
> I've broken the building of src/mingw with revision 784
> I have the fix here ready to commit, but my local trunk copy decided to
> get broken.
> I hope to be able to commit the fix in the next hours.
> You can svn up to revisi
Hi all,
I am committing this small patch to make lseek handle the case there an
invalid whence
is passed. The error handling is delegated to SetFilePointer.
(Remember that there is no errno in coredll.dll, only
Get|SetLastError(), which doesn't map perfectly).
Cheers,
Pedro Alves
---
Hi all,
Here are the libstdc++ cleanups possible because of the previous
libmingxex commits.
Cheers,
Pedro Alves
src/gcc/libstdc++-v3/ChangeLog.ce:
2006-11-12 Pedro Alves <[EMAIL PROTECTED]>
* config/io/basic_file_stdio.cc (__gnu_internal::read,
__gnu_internal:
MT);
#else
printf ("_M_ARMT not defined\n");
#endif
return 0;
}
Cheers,
Pedro Alves
---
src/gcc/gcc/ChangeLog.ce:
2006-11-12 Pedro Alves <[EMAIL PROTECTED]>
* config/arm/wince-pe.h (TARGET_OS_CPP_BUILTINS): Delete _M_ARM
and ARM.
---
src/w32api/ChangeLog.ce:
2006-1
lp32Snapshot is Windows CE only. Desktop versions of Windows
use the standard
CloseHandle to close a snapshot. That is described here:
http://msdn2.microsoft.com/en-us/library/ms682489.aspx
The committed patch is attached, and the updated ChangeLog entry is below.
Cheers,
Pedro Alves
ent ways other than
looking elsewhere.
Cheers,
Pedro Alves
---
-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IB
en-us/wceui40/html/cerefImageList_Copy.asp
ImageList_DrawIndirect:
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/wceui40/html/cerefImageList_DrawIndirect.asp
---
src/w32api/ChangeLog.ce:
2006-11-12 Pedro Alves <[EMAIL PROTECTED]>
* include/commctrl.h (IMAGELISTDRAWPAR
/wce50lrfSHGetShortcutTarget.asp
**
SHGetSpecialFolderPath:
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/wceui40/html/cerefSHGetSpecialFolderPath.asp
Cheers,
Pedro Alves
---
src/w32api/ChangeLog.ce:
2006-11-12 Pedro Alves <[EMAIL PROTECTED]>
* include/shel
tPoint
<http://msdn2.microsoft.com/en-us/library/ms901136.aspx> function."
I have then implemented and committed both *GetTextExtentPoint32 and
**GetTextExtentPointW in terms of
**GetTextExtentExPoint*.
Cheers,
Pedro Alves
---
src/w32api/ChangeLog.ce:
2006-11-12 Pedro Alves <[EMAIL
0, 0, NULL, DI_NORMAL*)*."
Cheers,
Pedro Alves
---
src/w32api/ChangeLog.ce:
2006-11-12 Pedro Alves <[EMAIL PROTECTED]>
* include/winuser.h (DrawIcon): Implement in terms of DrawIconEx.
Index: winuser.h
===
--- winu
positive it is correct, but I'm
not 100% sure.
If you have anything to add, please speak up.
Cheers,
Pedro Alves
2006-11-12 Pedro Alves <[EMAIL PROTECTED]>
* include/winuser.h (GetSystemMenu): Implement as macro.
Index
their names, you can guess what they do.
Concerning hiding the functions in w32api that are not implemented:
My current approach is to disable them as I stumble on them. There are
so many, that going through all of them, would
litter the code with #ifn
cluded, ARM will not be defined.
This matches the behaviour of the i386 gcc based compilers that use
w32api (at least Cygwin and MinGW).
Cheers,
Pedro Alves
-
Using Tomcat but need to do more? Need to support web services, sec
ng to build a
>pkfuncs.h file. If there is interest, I'll push it to cegcc.
>
>
Yes, please :)
Cheers,
Pedro Alves
-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done qu
Pedro Alves wrote:
Paul Sokolovsky wrote:
But still, liboldname-ce would rather not export non-existent
functions.
Oh, you are right, that was indeed a big thinko on my part.
There is indeed a thin write implementation in libmingwex.a, but you
are right in that liboldname should not
Oh, you are right, that was indeed a big thinko on my part.
There is indeed a thin write implementation in libmingwex.a, but you are
right in that liboldname should not
export those functions, since it maps the functions into coredll.dll.
I will work on a proper fix ASAP, thanks!
Cheers,
Ped
w32ce-g++ main.cpp -o main.exe
$ (no output)
Have you seen a real need to put those std c definitions there? If so,
we probably miss an include somewhere,
else, I'd rather remove them from winbase.h.
Cheers,
Pedro Alves
-
led?
Looks like something is trying to call the native (i386) as, but it ends up
picking that script on the way.
I guess building with CFLAGS=-v would help pin this down.
In the meantime, with a comment, that change is ok for trunk.
Cheers,
Pedro Alves
Danny Backx wrote:
> On Sat, 2006-11-18 at 09:53 +0000, Pedro Alves wrote:
>
>> Did you ever get to analise this more? Look at the path, and understand why
>> the i386 as is being called?
>> Looks like something is trying to call the native (i386) as, but it ends up
&
ult.asp?url=/library/en-us/wcekernl/html/ceconToolHelpStructures.asp
>
>
>
Looking at and the pages linked from it:
http://msdn2.microsoft.com/en-us/library/ms686735.aspx
Those fields are only present on Windows CE, not in desktop Windows, so when
applying I moved those fields inside #ifdef
Kevin O'Connor escreveu:
> On Thu, Nov 16, 2006 at 02:17:53PM +0000, Pedro Alves wrote:
>
>> Pedro Alves wrote:
>>
>>> Paul Sokolovsky wrote:
>>>
>>>
>>>> But still, liboldname-ce would rather not export non-existent
&
Danny Backx escreveu:
> On Sat, 2006-11-18 at 21:52 +0000, Pedro Alves wrote:
>
>>> It happens to do that because . is in my path before system directories;
>>> this may be why you don't see it :
>>> sh-3.00$ findpath as
>>>
in?
Could it be that we are emitting some arm insns that that pda's
processor doesn't support? We are defaulting
to armv4.
But (most probably) you are calling some function that is not supported
by WinCE 2 that then returns
a NULL of an invalid handle, that you are not checking?
Just gues
; implemented like this to begin with. Does anyone see a reason ?
>
>
Maybe ask the binutils mailing list?
Cheers,
Pedro Alves
-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay pane
Kevin O'Connor escreveu:
> On Mon, Nov 20, 2006 at 02:27:59PM +0000, Pedro Alves wrote:
>
>> Pedro Alves escreveu:
>>
>>> Kevin O'Connor escreveu:
>>>
>>>
>>>> On Mon, Nov 20, 2006 at 01:18:21AM +, Pedro Alv
Kevin O'Connor escreveu:
> On Thu, Nov 23, 2006 at 04:10:44AM +0000, Pedro Alves wrote:
>
>> Kevin O'Connor escreveu:
>>
>>> As to a fix - I'm a bit confused by the current mingw crt1.c code.
>>> Why do we have WinMainCRTStartup call
nged.
(Just take a look to the existing entries in mingw/ChangeLog, not
ChangeLog.mingw32ce, to get a feeling.)
When we will submit our code up for upstream inclusion, we will be asked for
ChangeLogs in GNU style, so we better use it too from the beginning.
Attached is the committed patch.
Cheers,
.
We where already using gcrt3.o, so this was a missing cleanup, with
precedents.
Cheers,
Pedro Alves
src/mingw/ChangeLog.mingw32ce:
2006-11-27 Pedro Alves <[EMAIL PROTECTED]>
* Makefile.in: Adjust so mingw32ce uses crt3.o and dllcrt3.o.
* dllcrt1.c: Remove extra space in c
This patch makes ld ignore symbols from crt3.o and dllcrt3.o,
as it is already doing for the other runtimes.
Cheers,
Pedro Alves
src/binutils/ld/ChangeLog.ce:
2006-11-27 Pedro Alves <[EMAIL PROTECTED]>
* pe-dll.c (autofilter_objlist) : Add crt3.o and dllcrt3.o.
Index: pe
This patch makes gcc use crt3.o and dllcrt3.o.
Cheers,
Pedro Alves
src/gcc/gcc/ChangeLog.ce:
2006-11-27 Pedro Alves <[EMAIL PROTECTED]>
* config/arm/mingw32.h (STARTFILE_SPEC): Add crt3.o and dllcrt3.o
instead of crt2.o and dllcrt2.o.
Index: config/arm/min
Hi all,
This patch makes the default DllMain call DisableThreadLibraryCalls.
There is nothing in
dllcrt1.c that depends on DLL_THREAD_ATTACH or DLL_THREAD_DETACH,
so I think it is pretty safe.
Does anyone see a problem?
Cheers,
Pedro Alves
src/mingw/ChangeLog.mingw32ce:
2006-11-27 Pedro
the Win32 API since it's
really part of Platform SDK.
*/
Something similar would be fine.
- Don't use c++ style comments (//).
Cheers,
Pedro Alves
-
Take Surveys. Earn Cash. Influence the Future of IT
J
hould fix it. Here is the ChangeLog entry:
src/mingw-fake_crt/ChangeLog:
2006-11-28 Pedro Alves <[EMAIL PROTECTED]>
* crt3.s: Rename from crt2.s.
* install.sh: Adjust to build crt3.o.
Cheers,
Pedro Alves
---
or.
I've tested it manually using -march=armv4, -march=armv5, etc, and with
-mcpu=arm8, -mcpu=xscale, etc.
I will commit it later if you confirm it works for you, or you can
commit it yourself if you want.
Cheers,
Pedro Alves
Inde
Pedro Alves escreveu:
>
> I've tested it manually using -march=armv4, -march=armv5, etc, and
> with -mcpu=arm8, -mcpu=xscale, etc.
I meant like this:
$touch test.c
$arm-wince-mingw32ce-gcc -mcpu=xscale -dD -E test.c | grep ARM
Cheer
).
>
> The reason I'm also asking about the installation path is I'd want to
> create an RPM based on these scripts. Do we keep the funky path for now
> (I used to do the same with my scripts in scripts/linux), or do we move
> to a "standard" place.
>
What d
Danny Backx wrote:
>On Mon, 2006-12-18 at 11:56 +0000, Pedro Alves wrote:
>
>
>>[EMAIL PROTECTED] wrote:
>>
>>
>>>Revision: 839
>>> http://svn.sourceforge.net/cegcc/?rev=839&view=rev
>>>Author: dannybackx
>>&g
Danny Backx wrote:
> On Mon, 2006-12-18 at 10:23 +0000, Pedro Alves wrote:
>
>>Danny Backx wrote:
>>
>>>The reason I'm also asking about the installation path is I'd want to
>>>create an RPM based on these scripts. Do we keep the funky path for no
Danny Backx escreveu:
> On Mon, 2006-12-18 at 17:15 +0000, Pedro Alves wrote:
>> I think cross-toolchains are
>> special, in the sense that they shouldn't be mixed with the
>> system headers/libs. I don't have an official pointer describing
>> it, but it seems
build_w3api_dir
$path_to_w32api_src/configure --host=arm-wince-mingw32ce
--prefix=$mingw32_prefix
Cheers,
Pedro Alves
---
2006-12-18 Pedro Alves <[EMAIL PROTECTED]>
* libce: New subdir.
(libce/CellCore.def, libce/Makefile.in, libce/VBarCall.def,
libce/VoiceCtl.def, libce/W
Hi all,
I just committed a change to remove the _M_ARM and ARM definitions from
w32api/headers. gcc now defines those are built-ins, so we don't need
them there anymore.
Cheers,
Pedro Alves
---
2006-12-18 Pedro Alves <[EMAIL PROTECTED]>
* include/winnt.h (_M_ARM, A
Pedro Alves escreveu:
> Hi all,
>
> This patch makes the default DllMain call DisableThreadLibraryCalls.
> There is nothing in
> dllcrt1.c that depends on DLL_THREAD_ATTACH or DLL_THREAD_DETACH,
> so I think it is pretty safe.
>
> Does anyone see a problem?
>
>
ge like
"no include path to look for errno.h" or whatever.
What this means in the end is that you have to look in the gcov sources
for #include instances, and wrap them in '#ifdef HAVE_ERRNO_H'
and possibly add an AC_CHECK_HEADER for errno.h to to gcov's
configure.in/configu
d be that you set a Registry key,
instead of doing a shell export. Much more flexible, IMHO.
I would go as far as implementing the a function that reads from
the registry with the same interface as getenv, returning a pointer
into a local static buffer, to keep our patch isolated, and call th
of doing a shell export. Much more flexible, IMHO.
>>I would go as far as implementing the a function that reads from
>>the registry with the same interface as getenv, returning a pointer
>>into a local static buffer, to keep our patch isolated, and call that
>>instead of g
Danny Backx wrote:
> On Fri, 2006-12-29 at 15:35 +0000, Pedro Alves wrote:
>
>>Danny Backx wrote:
>>
>>
>>>>>@@ -1,4 +1,10 @@
>>>>> #
>>>>>+# This is a Makefile.in for very limited use : only create and install
>>>>
Danny Backx wrote:
> Second set of patches included.
>
> My option would be to get this in SVN, then continue work on the getenv
> replacement and command line option.
>
Ok, go ahead.
Cheers,
Pedro Alves
Danny Backx escreveu:
> On Sat, 2006-12-30 at 12:45 +0100, Danny Backx wrote:
>> On Sat, 2006-12-30 at 10:38 +0000, Pedro Alves wrote:
>>> Danny Backx wrote:
>>>> Second set of patches included.
>>>>
>>>> My option would be to get this in SVN,
,$(target_alias)))
inst_bindir:=$(tooldir)/bin
inst_includedir:=$(tooldir)/include
inst_libdir:=$(tooldir)/lib
inst_docdir:=$(prefix)/share/doc/mingw-runtime
else
Should work here too, and when/if we have a top level profile/configure,
I think
ver port. It is already working
much better than
our current gdb, but there is this one annoying bug related to dwarf
debugging and passing
big structures on the stack. Me goes to have dinner and nice bottle of
red wine for inspiration :)
Cheers,
Pedro Alves
his case it justifies an
extra dependency. This is the same rationale I used to introduce
src/gcc/libstdc++-v3/config/os/mingw32ce/runtimeopts.h, and replaced
every getenv call in libstdc++ with runtime_opts::force_new_p().
Cheers,
Pedro Alves
---
buildable) test case would be great. If we
can't reproduce the problem it is very hard to come
up with a fix. Try reducing as much as you can while still being able
to see the problem. That usually helps in finding the culprit itself.
(snipping Portuguese translation, sin
1 - 100 of 534 matches
Mail list logo
