--
Best regards,
LH_Mouse
From a53012e455c7ba781a3fedb8d0252878197aeb51 Mon Sep 17 00:00:00 2001
From: Liu Hao
Date: Wed, 27 Jan 2021 10:50:13 +0800
Subject: [PATCH 4/4] include/crt: Ensure declarations of `_*_nolock` match
MSVCRT version
Signed-off-by: Liu Hao
---
mingw-w64-headers/crt
--
Best regards,
LH_Mouse
From 7cc6b288d5be31ac772d61e390a4b9d826667d29 Mon Sep 17 00:00:00 2001
From: Liu Hao
Date: Wed, 27 Jan 2021 10:50:13 +0800
Subject: [PATCH 1/4] include/crt: Ensure declarations of `*wcsym*` match
MSVCRT version
Signed-off-by: Liu Hao
---
mingw-w64-headers/crt
--
Best regards,
LH_Mouse
From 44a282f808eefd14f1fe2f6b410e489d0ccf90fb Mon Sep 17 00:00:00 2001
From: Liu Hao
Date: Wed, 27 Jan 2021 11:20:33 +0800
Subject: [PATCH 2/4] crt/msvcr90: Add aliases for `fgetc_nolock` and
`getwc_nolock`
Signed-off-by: Liu Hao
---
mingw-w64-crt/lib32/msvcr90
在 2021/1/27 上午4:02, Kacvinsky, Tom 写道:
>
> I am using a stock MSYS2 + MinGW-w64 package for GCC. So if there is a
> misconfiguration,
> it isn't from me. All I am doing is including cwhar. The thing is, our
> product parses the
> header files in a TU and "instruments" them, so they end up in
在 2021/1/25 下午10:40, Kacvinsky, Tom 写道:
> At the bottom you'll find the link errors I am getting
>
>
`iscsym` requires MSVCR80 or UCRT, and is not available in MSVCRT. So looks
like it is a
misconfiguration.
`strncat_s_l` on the other hand should be be an inline function, but it is
在 2021-01-25 22:25, Kacvinsky, Tom 写道:
Hi,
Attached is a file of the link errors I am getting. There are bunch of them.
I am using the
built-in specs, so the run time library I am working with is libmsvcrt.a
Please attach text files as .TXT files, or paste their contents inline;
在 2021-01-24 02:05, Hannes Domani via Mingw-w64-public 写道:
Am Samstag, 23. Januar 2021, 16:46:18 MEZ hat Jeroen Ooms
Folgendes geschrieben:
A user of the R programming language has reported that std::regex
causes a hang for certain regular expressions when running in Japanese
locale. I was
This was done with the following commands:
sed -Ei 's/^_(str|wcs)lwr\>.*$/&\n\1lwr == _\1lwr/' `grep -Elr '\<_strlwr\>'
mingw-w64-crt`
git checkout mingw-w64-crt/lib-common/api-ms-win-crt-string-l1-1-0.def
Signed-off-by: Liu Hao
---
mingw-w64-crt/lib-common/msvcr120_a
在 2021/1/20 上午5:44, Martin Storsjö 写道:
> Signed-off-by: Martin Storsjö
> ---
> Added memory clobbers for YieldProcessor() too - now actually amended
> the patch before sending...
> ---
> mingw-w64-headers/crt/_mingw.h.in | 8
> mingw-w64-headers/include/winnt.h | 6 ++
> 2 files
05b0bb43124b041da360ba9adcbaab8430be6d18 Mon Sep 17 00:00:00 2001
From: Liu Hao
Date: Wed, 6 May 2020 21:49:18 +0800
Subject: [PATCH] libgomp: Don't hard-code MS printf attributes
---
libgomp/libgomp.h | 13 ++---
1 file changed, 10 insertions(+), 3 deletions(-)
diff --git a/libgomp
在 2021/1/19 上午11:54, Zebediah Figura (she/her) 写道:
>
> Do you mean that the writer should change the above code to e.g.
> "__atomic_load_n(, ___ATOMIC_ACQUIRE)", or that the read already
> implies acquire semantics? I didn't think that the latter was true, but
> I don't trust that judgement;
在 2021/1/19 上午3:32, Zebediah Figura (she/her) 写道:
> I guess I ask because it's not clear to me whether it should be
> necessary either. It doesn't seem to be present for x86, but there might
> be code out there that does something like
>
> "while (!var) YieldProcessor();"
>
> and expects it to
在 2021-01-18 22:11, Martin Storsjö 写道:
LLD only got support for mingw targets in LLD 6 anyway, so this only affects LLD 6 and 7, and usage
of LLD in mingw contexts was very scarce at the time (only in llvm-mingw at the time, afaik).
Nowadays msys2 also has picked it up, but both llvm-mingw and
在 2021-01-18 04:16, Martin Storsjö 写道:
This reverts commit ccda8b5229e428a6fa7036140778b23d11b554e8 and
removes the code for providing __CTOR_LIST__ here.
This was only needed for LLD 7 and older, while LLD 8 got support for
providing __CTOR_LIST__ by the linker, just as GNU ld does.
This gets
在 2021-01-18 04:05, Martin Storsjö 写道:
Signed-off-by: Martin Storsjö
---
Added memory clobbers for MemoryBarrier().
---
mingw-w64-headers/crt/_mingw.h.in | 8
mingw-w64-headers/include/winnt.h | 6 ++
2 files changed, 14 insertions(+)
This patch looks good to me.
--
Best
在 2021/1/16 上午6:42, Alex Henrie 写道:
> On Thu, Dec 17, 2020 at 7:47 AM Liu Hao wrote:
>>
>> ReactOS uses the name `Reserved` [1] so I wonder how much existent code
>> expects this one.
>>
>> (BTW they selected a different name for the 5th field.)
>
> React
在 2020/12/26 上午4:41, Martin Storsjö 写道:
>
> Maybe they don't do that right in any places right now, but despite that, I
> believe there's a value
> in distinguishing between __MSVCRT_VERSION__ 0xA00 (aka msvcr100.dll) and a
> modern msvcrt.dll on
> Windows 7, so something like
在 2021/1/14 上午5:14, Martin Storsjö 写道:
>
> --- a/mingw-w64-headers/crt/_mingw.h.in
> +++ b/mingw-w64-headers/crt/_mingw.h.in
> (...)
> + __asm__ __volatile__("unimplemented");
I suggest an `#error` here so the error could be caught earlier.
> --- a/mingw-w64-headers/include/winnt.h
> +++
在 2021-01-13 10:54, Clinton Reddekop 写道:
Hello,
I have mingw-w64 compiler tools version 8.1.0, downloaded here:
https://sourceforge.net/projects/mingw-w64/files/Toolchains targetting
Win32/Personal Builds/mingw-builds/installer/mingw-w64-install.exe/download
niXman [1] used to maintain those
在 2021-01-12 01:37, Biswapriyo Nath 写道:
Hmmm why aren't those IDLs imported from Wine?
1. Huge confusion. IExtractIcon is in -- shlobj_core.h in WinSDK,
shlobjidl.idl in wine, shlobj.h in mingw-w64.
2. Files differ largely between mingw-w64 and wine.
This thing was added in WinSDK 8.1 and
在 2020-12-19 06:50, Jeremy Drake via Mingw-w64-public 写道:
Have there been any further ideas on this? I am concerned the current
state of affairs, accessing freed memory during tls destruction of objects
in the .exe module, could cause crashes in objects more complicated than
simple logging
在 2021/1/11 下午11:00, Stephen Kitt 写道:
> On Mon, 11 Jan 2021 22:07:42 +0800, Liu Hao wrote:
> [...]
>> There is no special meaning about `w64` itself. `mingw32` on the other hand
>> specifies the ABI, so all `i686-*-mingw32` targets are considered
>> ABI-compatible.
>
在 2021-01-11 22:43, LRN 写道:
On 11.01.2021 17:07, Liu Hao wrote:
You invoked `i686-w64-mingw32-gcc` on your machine (the 'host'), which produced
a.exe suitable for
Windows on i686 (the 'target'). `x86_64-w64-mingw32-gcc` is similar.
Actually, if you want to be pedantic, his machine is 'build
在 2021-01-08 13:26, Biswapriyo Nath 写道:
So, I came up with this bizarre idea. I copy only the IExtractIcon{A,
W} interfaces from Wine's shlobjidl.idl file and create a dummy
shobj.idl file in mingw-w64. Then I generate the header file and
replace the generated interface to our shlobj.h file
在 2021-01-10 23:46, Peng Yu 写道:
Welcome to the world of GNU. These are cpu-vendor-os triplets (yes, mingw32 is
not an OS, welcome to the club) that identify the toolchain. i686 is 32-bit,
x86_64 is 64-bit - that's all you need to now with regards to mingw. A bit more
info can be found on the
在 2021-01-05 13:25, Biswapriyo Nath 写道:
Thanks. These patches look good to me. Pushed.
--
Best regards,
LH_Mouse
OpenPGP_signature
Description: OpenPGP digital signature
___
Mingw-w64-public mailing list
Mingw-w64-public@lists.sourceforge.net
在 2021/1/3 上午12:01, JonY via Mingw-w64-public 写道:
> On 1/2/21 3:08 PM, Biswapriyo Nath wrote:
>> Any reply on this patch?
>>
>>
>
> I don't really have any opinions of this patch, I'll apply both if there are
> still no objections in
> a few days.
>
>
>
LGTM.
BTW I will receive my new
在 2020/12/27 上午5:44, Martin Storsjö 写道:
>
> I'm not entierely sure I agree with this - I think different ecosystems have
> different expectations
> here. (It'd be interesting to e.g. rebuild all of msys2's mingw package
> repository with such
> headers, and see how many of them fail to run on
在 2020/12/24 上午11:02, Liu Hao 写道:
>
> I attached a small test program for this.
>
>
Here it is. Please remove all `.txt` suffixes from file names.
--
Best regards,
LH_Mouse
#!/bin/sh
windres lconv.rc -o lconv.rc.o
x86_64-w64-mingw32-gcc -std=c99 lconv.c lconv.rc.o -lmsv
在 2020/12/24 上午1:07, Christoph Reiter 写道:
> On Wed, Dec 23, 2020 at 6:35 AM Liu Hao wrote:
>> These fields seem to have been added since MSVCR100, so please add a
>> necessary check for
>> `__MSVCRT_VERSION__`.
>>
>> This also means that CPython has to def
在 2020/12/23 22:17, Biswapriyo Nath 写道:
>> $ widl -DBOOL=WINBOOL -h -o axextend.{h,idl}
>> axextend.idl:79: error: syntax error, unexpected aIDENTIFIER
>
> axextend.idl is not meant to be generated. axextend.idl is included in
> stmif.idl and strmif.idl has to be generated.
>
>
Ah thanks for
在 2020/12/23 0:15, Biswapriyo Nath 写道:
> Subject: [PATCH] headers: Move some interfaces declaration to axextend.idl
> from strmif.idl.
>
> This helps to generate proper header files instead of using manual cpp_quote.
>
> Signed-off-by: Biswapriyo Nath
> ---
>
在 2020/12/23 上午6:22, Christoph Reiter 写道:
> diff --git a/mingw-w64-headers/crt/locale.h b/mingw-w64-headers/crt/locale.h
> index 068f456c..4f783a91 100644
> --- a/mingw-w64-headers/crt/locale.h
> +++ b/mingw-w64-headers/crt/locale.h
> @@ -61,6 +61,14 @@ extern "C" {
> char n_sep_by_space;
>
在 2020/12/22 15:20, Biswapriyo Nath 写道:
> Subject: [PATCH] headers: Add more stub macos in sal.h.
>
> Required for assimp https://github.com/assimp/assimp.git
>
> Signed-off-by: Biswapriyo Nath
> ---
> mingw-w64-headers/include/sal.h | 121
> 1 file changed,
在 2020/12/22 下午12:52, Biswapriyo Nath 写道:
>> so shouldn't there be an `#if _WIN32_WINNT
>
> WinSDK does not have the winver guard there. If we have to apply this
> rule we have to edit many other headers (if not hundreds). Updated
> patch attached as asked.
>
I still consider it better to have
在 2020/12/22 上午1:59, Biswapriyo Nath 写道:
> diff --git a/mingw-w64-headers/include/handleapi.h
> b/mingw-w64-headers/include/handleapi.h
> index 0e0fd82..6f122b3 100644
> --- a/mingw-w64-headers/include/handleapi.h
> +++ b/mingw-w64-headers/include/handleapi.h
> @@ -18,6 +18,7 @@ extern "C" {
>
在 2020/12/21 17:11, Biswapriyo Nath 写道:
> Any update on this?
>
>
Apologies for missing this. This patch looks good to me and I pushed it.
--
Best regards,
LH_Mouse
signature.asc
Description: OpenPGP digital signature
___
Mingw-w64-public
在 2020/12/21 3:19, Biswapriyo Nath 写道:
> Subject: [PATCH] headers: Add mmddk.h file.
>
> Required for libdshowcapture https://github.com/obsproject/libdshowcapture.git
>
> Signed-off-by: Biswapriyo Nath
> ---
> mingw-w64-headers/include/mmddk.h | 544 ++
> 1 file
在 2020/12/20 2:22, Jacek Caban 写道:
>
>
> The patch looks good to me.
>
>
Thanks; I pushed this one.
--
Best regards,
LH_Mouse
signature.asc
Description: OpenPGP digital signature
___
Mingw-w64-public mailing list
在 2020/12/19 下午2:53, Biswapriyo Nath 写道:
> Can these be implemented with inline assembly (rol, ror)? That may be
> one liner. This 8, 16 treatment is also needed for
> _InterlockedExchange, _InterlockedOr etc. intrinsics.
>
>
>
Personally I think we had better avoid assembly for two reasons:
在 2020/12/18 下午5:16, Biswapriyo Nath 写道:
>> Why did you remove those alignment attributes?
>
> The way you have explained, I thought aligning structure is not
> allowed. My mistake! Here is update #3.
>
>
>
Thanks. I pushed this one.
Side note: AFAICT all those explicitly defaulted
在 2020/12/17 下午11:40, Biswapriyo Nath 写道:
> Update #1.
>
>
Why did you remove those alignment attributes? `XMFLOAT4A` etc. have to be
aligned (note the final
`A` in its name) so they must be declared with those attributes.
--
Best regards,
LH_Mouse
signature.asc
Description: OpenPGP
在 2020/12/17 22:28, Biswapriyo Nath 写道:
> The SYSTEM_PROCESSOR_INFORMATION structure is not even in Microsoft's
> Windows SDK or in documentation. So, changing "unknown" to correct one
> (from combase.pdb) may not cost anything ;)
>
>
>
ReactOS uses the name `Reserved` [1] so I wonder how much
在 2020/12/17 3:08, Biswapriyo Nath 写道:
> +#define DIRECTX_MATH_VERSION 314
> +
> +#define XM_CONST const
> +#define XM_CONSTEXPR
This can be defined as:
```c++
#if __cplusplus >= 201103L
# define XM_CONSTEXPR constexpr
#else
# define XM_CONSTEXPR
#endif
```
> +__declspec(align(16)) struct
在 2020/12/17 10:09, Alex Henrie 写道:
> Signed-off-by: Alex Henrie
> ---
> This field is the current number of processors in the system plus the
> number of processors that could be added without rebooting.
>
> https://www.geoffchappell.com/studies/windows/km/ntoskrnl/api/ex/sysinfo/processor.htm
在 2020/12/16 2:20, Biswapriyo Nath 写道:
> Subject: [PATCH 4/4] headers: Add STATIC_CODECAPI macros in codecapi.h.
>
> Signed-off-by: Biswapriyo Nath
> ---
> mingw-w64-headers/include/codecapi.h | 311 +++
> 1 file changed, 311 insertions(+)
>
These four patches look
在 2020/12/16 6:18, Clinton Reddekop 写道:
> Hello,
> What is the correct way to use ncurses in my programs compiled with
> mingw-w64?
>
> I use this version of the compiler on Windows 10: gcc
> (x86_64-posix-seh-rev0, Built by MinGW-W64 project) 8.1.0
>
> The compiler came with some ncurses stuff
在 2020/12/15 14:36, Biswapriyo Nath 写道:
> Subject: [PATCH] headers: Add IAudioClient3 and IAudioAmbisonicsControl in
> AudioClient.idl.
>
> Signed-off-by: Biswapriyo Nath
> ---
> mingw-w64-headers/include/audioclient.idl | 51 ++-
> 1 file changed, 50 insertions(+), 1
在 2020/12/15 19:57, Jacek Caban 写道:
> Signed-off-by: Jacek Caban
> ---
> mingw-w64-headers/Makefile.am | 4
> 1 file changed, 4 insertions(+)
>
>
>
These patches look good to me.
--
Best regards,
LH_Mouse
signature.asc
Description: OpenPGP digital signature
/msys2/MINGW-packages/issues/7437
Signed-off-by: Liu Hao
---
.../include/psdk_inc/intrin-impl.h| 48 +++
1 file changed, 48 insertions(+)
diff --git a/mingw-w64-headers/include/psdk_inc/intrin-impl.h
b/mingw-w64-headers/include/psdk_inc/intrin-impl.h
index 072a98f6..27
在 2020/12/11 18:26, Christer Solskogen 写道:
> While compiling the mingw-w64 cross compiler I got this with gcc-10.2.0:
>
> (... snip ...)
>
> Is this a gcc problem or mingw-w64 problem? Or perhaps "my" problem if I've
> configured something
> wrongly.
> mingw-w64 is from the latest v8.x branch.
在 2020/12/9 下午11:39, Biswapriyo Nath 写道:
> Updated #1.
>
Thanks. This patch looks good. Pushed.
--
Best regards,
LH_Mouse
signature.asc
Description: OpenPGP digital signature
___
Mingw-w64-public mailing list
在 2020/12/8 16:51, Biswapriyo Nath 写道:
> d3d9 headers are imported from wine. So, you have to send any changes
> to the wine mailing list first. Also in the patch, you have removed
> two defines which are used in wine source code. Those two manifest
> constants are not present in current Microsoft
在 2020/12/7 上午5:30, Martin Storsjö 写道:
>
> My thought here was that previously by including malloc.h, you got a working
> _mm_malloc and _mm_free
> (defined by pointing at other functions), so there could be code relying on
> getting these functions
> visible by just relying on malloc.h. (The
在 2020/12/4 下午3:44, Martin Storsjö 写道:
> Instead just remove the _mm_malloc redefinition and include the
> compiler's mm_malloc.h header. This makes sure that regardless of
> header include order, calls to _aligned_malloc and _mm_malloc will
> always end up to the same function, avoiding risks of
在 2020/12/2 上午2:24, Biswapriyo Nath 写道:
> Some projects like apitrace and assimp need DirectXMath header files.
> Is it possible to add those here? I am asking this because those
> header files contain many MSVC specific SIMD intrinsics and
> preprocessors.
>
>
>
Are you referring to those
在 2020/11/26 21:03, Biswapriyo Nath 写道:
> Subject: [PATCH 2/2] headers: Add missing APIs in winnls.h.
>
> Signed-off-by: Biswapriyo Nath
> ---
> mingw-w64-headers/include/winnls.h | 55 +-
> 1 file changed, 54 insertions(+), 1 deletion(-)
Thanks. These patches look
在 2020/11/26 12:03, Biswapriyo Nath 写道:
> A snippet of that code will be helpful. In Windows SDK, winver.h does
> not have that winapifamily.h include. I think winver.h should not be
> directly included in source code. winres.h is generally included in
> resource files (.rc). Also each function in
在 2020/11/17 下午9:05, Martin Storsjö 写道:
> This function has existed in import libs and in special cased helper
> functions with GetProcAddress for a decade, but has been missing
> from public headers.
>
> Signed-off-by: Martin Storsjö
> ---
> mingw-w64-headers/crt/locale.h | 2 ++
> 1 file
在 2020/11/17 下午9:43, Martin Storsjö 写道:
> Remove arch specific ifdefs and instead add dummy symbols to the
> arm specific import libs.
>
> Also add a dummy symbol for __winitenv, fixing building apps
> that use a wmain() function for arm/aarch64 when targeting msvcrt.dll.
>
> Signed-off-by:
在 2020/11/16 17:52, gmail Vladimir Koković 写道:
> I can't believe it but it's true that I can't do clean/build for libodbc++
> because my linux mingw
> (gcc 10.2) is badly packaged!
> I checked against http://winlibs.com/ gcc 10.2 and it is evident that in my
> linux mingw libstdc is
> missing
在 2020/11/16 下午1:57, gmail Vladimir Koković 写道:
> Nothing has changed since the destructor was added, so the results are the
> same.
> Although the try/catch block is active vkSQLException is not caught:
>
Did you rebuild libodbc++ with exact the same version of GCC that you used to
compile
在 2020/11/16 上午5:54, gmail Vladimir Koković 写道:
> Hi,
> Let me try to describe what kind of problem I have with mingw-w64,
>
> because it becomes a nightmare for me, despite the fact that I have over 30
> years of experience with
> C and c ++ compilers:
> My ERP consists of a main program and
在 2020/11/13 20:29, Biswapriyo Nath 写道:
> Subject: [PATCH 2/2] crt/libsrc: Include directmanipulation.h in uuid library.
>
> Signed-off-by: Biswapriyo Nath
> ---
> mingw-w64-crt/libsrc/uuid.c | 1 +
> 1 file changed, 1 insertion(+)
Thanks. These patches look good. Pushed.
--
Best regards,
在 2020/11/13 下午5:44, Martin Storsjö 写道:
> These routines are used on msvcrt.dll on arm, which only provides
> log(). For exact powers of two, produce an exact result instead
> of one that only is numerically close.
>
> Defer to the normal implementation for anything other than the
> simple cases
在 2020/11/13 下午5:04, Martin Storsjö 写道:
> These routines are used on msvcrt.dll on arm, which only provides
> log(). For exact powers of two, produce an exact result instead
> of one that only is numerically close.
>
> Defer to the normal implementation for anything other than the
> simple cases
在 2020/11/12 14:20, Biswapriyo Nath 写道:
> Subject: [PATCH] headers: Replace BOOL with WINBOOL in audioclient.idl.
>
> Signed-off-by: Biswapriyo Nath
> ---
> mingw-w64-headers/include/audioclient.idl | 10 +-
> 1 file changed, 5 insertions(+), 5 deletions(-)
Thanks. This patch looks
在 2020/11/8 下午9:43, Biswapriyo Nath 写道:
> Subject: [PATCH 2/2] headers/d2d1_1: Add macros for C style interfaces.
>
> Signed-off-by: Biswapriyo Nath
> ---
> mingw-w64-headers/include/d2d1_1.h | 124 +
> 1 file changed, 124 insertions(+)
Thanks. Both patches look
在 2020/11/8 6:28, Nikolay Sivov 写道:
> Signed-off-by: Nikolay Sivov
> ---
> mingw-w64-crt/lib32/mf.def | 1 +
> 1 file changed, 1 insertion(+)
>
Thanks. I pushed this one to master too.
--
Best regards,
LH_Mouse
___
Mingw-w64-public mailing
在 2020/11/7 23:45, Biswapriyo Nath 写道:
> From ddfb9b99bd24ff9dbfb028fa3e3f9a3b721213c0 Mon Sep 17 00:00:00 2001
> From: Biswapriyo Nath
> Date: Sat, 7 Nov 2020 20:23:49 +0530
> Subject: [PATCH] crt/dxguid: Add dxgi headers.
>
> Signed-off-by: Biswapriyo Nath
> ---
>
在 2020/11/5 21:37, Vincent Torri 写道:
>
> i can add what is missing part by part in d2d1.h. What is the standard
> procedure ? if there is a github mirror and if PR are a possible way
> to provide patches, i can do that
>
> regards
>
>
You may clone the mingw-w64 repo, commit your changes
在 2020/11/5 下午4:11, gmail Vladimir Koković 写道:
> Hi,
>
> After a sleepless night the morning wit is something completely different, so
> the analysis is now
> also different!
>
> I made a new simpler test example and the situation is as follows:
> 1. .cpp and .ii have two successive code blocks
[Please 'reply to list' or 'reply to all' instead of sending to me personally.]
在 2020/11/5 下午4:38, gmail Vladimir Koković 写道:
> HI,
>
> You didn't notice that _catchVK is with noexcept (false).
>
>
That's irrelevant. The compiler may assume that the `catch` block is never
executed, only
在 2020/11/5 上午6:13, gmail Vladimir Koković 写道:
> At this point I don't know if it's a gcc or ld problem, but I think I've
> found the real reason.
> If you look at ht and objdump dissassembler output, you can see that the exe
> file is missing the
> code for "__try1 (_catchVK)"!
>
>
Unlike on
在 2020/10/30 上午1:10, Jeremy Drake via Mingw-w64-public 写道:
> On Thu, 29 Oct 2020, Liu Hao wrote:
>
>> As far as I can tell it is only required by the c++ standard that
>> thread_local objects are destroyed in the reverse order that they have
>> been constructed and prior
在 2020/10/29 上午4:31, Martin Storsjö 写道:
>
> So with that in mind, winpthread _could_ flip the order it runs the
> destructors, and it would work
> with the libsupc++ implementation of __cxa_thread_atexit. Then winpthreads
> would run dtors in a
> different order than both glibc and darwin's,
在 2020/10/28 17:24, Martin Storsjö 写道:
> For these cases, the TLS destructors are normally executed by
> the callback set up by _register_thread_local_exe_atexit_callback,
> but if this is bypassed (if exiting via _exit or ExitProcess),
> skip these callbacks.
>
> This differs from what MSVC
在 2020/10/28 下午4:15, Martin Storsjö 写道:
>
> Regarding not running TLS dtors for the main thread when exiting with
> _exit(), this can be fixed
> easily with a change that looks like this:
> (... snip ...)
> */
> - run_dtor_list(_dtors);
> if (__mingw_module_is_dll) {
> +
在 2020/10/28 下午1:14, Jeremy Drake 写道:
> On Wed, 28 Oct 2020, Liu Hao wrote:
>
>> The first sentence of your commit message was 'the TLS callback function for
>> DLL_THREAD_DETACH is called after emutls has already cleaned up for the
>> thread'. As it is clearly a bug, a m
在 2020/10/28 下午12:18, Jeremy Drake 写道:
> On Wed, 28 Oct 2020, Liu Hao wrote:
>
>> I would like to put this patch on hold for a while because of two reasons:
>
> I agree that the destructor behavior is wrong at process exit, but it was
> before this patch too. I believe thi
在 2020/10/28 上午10:27, Liu Hao 写道:
>
> I would like to put this patch on hold for a while because of two reasons:
>
> 0) As far as I can tell this patch attempts to resolve this issue by storing
>a pointer to emutls data in a TLS slot, so it will eventually outlast
>
在 2020/10/28 上午4:55, Jeremy Drake via Mingw-w64-public 写道:
> The TLS callback function for DLL_THREAD_DETACH is called after emutls
> has already cleaned up for the thread, so the tls_dtors were always NULL
> at that time. Work around this by using Windows native TlsAlloc and
> friends.
>
>
在 2020/10/18 9:08, Alex Henrie 写道:
> Signed-off-by: Alex Henrie
> ---
> Needed by
> https://hg.mozilla.org/mozilla-central/file/88834ed9b3fc6e6879b1a7f1d4b987ff11ea4b99/mozglue/misc/WindowsProcessMitigations.cpp#l18
> ---
> mingw-w64-headers/include/winnt.h | 60 +++
>
在 2020/10/17 17:56, Kacvinsky, Tom 写道:
> This bombs out (i686 version)
>
> vapkay@vapkay5117nbj MINGW32
> /c/cygwin64/home/vapkay/vector/build/vc20__89949_update_mingw/vc/examples/environments/tutorial_ada
> $ gprbuild manager.gpr
> Unexpected exception while parsing knowledge base: raised
在 2020/10/14 上午2:57, Jacek Caban 写道:
>
> Signed-off-by: Jacek Caban
> ---
>
> This patch should fix the problem discussed in the other thread. It also came
> out recently in Mozilla:
> https://bugzilla.mozilla.org/show_bug.cgi?id=1667423
>
> mingw-w64-headers/include/dcomp.h | 14
在 2020/10/13 上午2:28, Biswapriyo Nath 写道:
> * Requires pthread_attr_getstack and pthread_attr_setstack in
> winpthreads for libwally-core[1] project.
> * Steps to build and test: Requires clang for test_clear in this project.
>
> cd libwally-core
> ./tools/autogen.sh
> mkdir bin; cd bin
> export
在 2020/10/12 上午1:50, Biswapriyo Nath 写道:
>
>
>
Thanks. Pushed.
--
Best regards,
LH_Mouse
signature.asc
Description: OpenPGP digital signature
___
Mingw-w64-public mailing list
Mingw-w64-public@lists.sourceforge.net
在 2020/10/10 下午2:25, sotrdg sotrdg 写道:
> D:/msys64/mingw64/x86_64-w64-mingw32/include/rpcndr.h:63:25: note:
> 'typedef unsigned char byte'
>63 | typedef unsigned char byte;
> | ^~~~
>
It's like that you have `using namespace std` somewhere in
在 2020/10/10 下午2:50, Biswapriyo Nath 写道:
>
>
>
>
Thanks. This patch looks good. Pushed.
--
Best regards,
LH_Mouse
signature.asc
Description: OpenPGP digital signature
___
Mingw-w64-public mailing list
Mingw-w64-public@lists.sourceforge.net
在 2020/10/8 3:05, Andrew Eikum 写道:
> Signed-off-by: Andrew Eikum
> ---
>
Next time, please do not commit 'Makefile.in' with other files. It's meant to
be auto-generated and can't be reviewed.
The other parts look good to me. I stripped this patch and pushed it. Thanks
for the work.
--
在 2020/10/8 4:30, Biswapriyo Nath 写道:
>
>
>
>
Thanks; these patches look good. Pusehd.
--
Best regards,
LH_Mouse
signature.asc
Description: OpenPGP digital signature
___
Mingw-w64-public mailing list
Mingw-w64-public@lists.sourceforge.net
在 2020/10/6 上午8:24, JonY via Mingw-w64-public 写道:
> On 10/5/20 6:44 PM, Biswapriyo Nath wrote:
>>
>>
>
> Looks good to me.
>
>
Thanks. I pushed these two patches to master.
--
Best regards,
LH_Mouse
signature.asc
Description: OpenPGP digital signature
在 2020/10/5 21:25, Biswapriyo Nath 写道:
> Thanks for the pedantic hint ^_^
>
>
>
Thanks. This series of patches look good to me. Pushed.
--
Best regards,
LH_Mouse
signature.asc
Description: OpenPGP digital signature
___
Mingw-w64-public mailing
在 2020/10/4 下午10:11, Biswapriyo Nath 写道:
> +#ifdef __cplusplus
> +};
> +#endif
> +
This semicolon follows the `extern "C"` block and has no effect. Should it be
removed?
I have noticed that this superfluous semicolon exists in MS headers as well.
They cause warnings with GCC in pedantic mode.
在 2020/10/3 13:32, Biswapriyo Nath 写道:
>
>
>
Thanks for these two patches (the other was about dxgi). I pushed them.
--
Best regards,
LH_Mouse
signature.asc
Description: OpenPGP digital signature
___
Mingw-w64-public mailing list
在 2020/9/26 12:58, Biswapriyo Nath 写道:
> Signed-off-by: Biswapriyo Nath
> ---
> mingw-w64-headers/include/winres.h | 5 +
> 1 file changed, 5 insertions(+)
>
Thanks for the patch. I pushed this one to master.
--
Best regards,
LH_Mouse
signature.asc
Description: OpenPGP digital
在 2020/9/20 19:00, Jean Pierre Cimalando via Mingw-w64-public 写道:
> Dear list,
>
> Hi I'm encountering this issue following the 8.0.0 update which
> included major directX API updates from the IDL files.
>
> I am finding that the COM APIs which are overloaded are renamed with
> underscore
在 2020/9/18 下午4:26, Biswapriyo Nath 写道:
> Try with adding "-D__USE_MINGW_ANSI_STDIO=1" in CFLAGS and/or CXXFLAGS.
>
>
That probably will not work, as it is the (now default) definition of it that
causes the issue, not the absence of it.
Patching GOMP source seems unavoidable [1].
As recent GCC
在 2020/9/18 上午6:25, Vicente Ferrari 写道:
> Hello,
>
> I've spent the past few days porting a DirectX program from the MSVC
> toolchain to MinGW-w64. The only big hurdle I
> encountered was the lack of popular API functions from the d3dcompiler.h
> header.
>
> In particular I found
在 2020/9/18 上午12:18, Ave Milia via Mingw-w64-public 写道:
> Hello,
>
> I am struggling to build mingw-w64-gcc on Manjaro Linux 20.1 with sjlj
> exceptions. This distribution uses very simple Arch Linux package rules by
> default [0], my modified version of the rules is here [1].
>
> (... snip
901 - 1000 of 1807 matches
Mail list logo
