Subject: [PATCH] OS/2: Try to commit memory above 1GB
Signed-off-by: Dave Yeo
---
configure | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/configure b/configure
index 9cd4fe6b..53dd809b 100755
--- a/configure
+++ b/configure
@@ -3964,7 +3964,7 @@ case $target_os in
On 02/04/16 03:50 AM, Vittorio Giovara wrote:
On Tue, Jan 26, 2016 at 7:16 PM, Luca Barbato wrote:
>On 23/01/16 19:35, Luca Barbato wrote:
>>And make sure to enable all the components needed
>>---
>> configure | 11 ---
>> 1 file changed, 8 insertions(+), 3 deletions(-)
>>
>
>Ping.
On 01/23/16 10:35 AM, Luca Barbato wrote:
And make sure to enable all the components needed
---
configure | 11 ---
1 file changed, 8 insertions(+), 3 deletions(-)
diff --git a/configure b/configure
index c5bcb78..675c112 100755
--- a/configure
+++ b/configure
@@ -489,7 +489,7 @@ enab
fixes assembling on OS/2
Signed-off-by: Dave Yeo
---
libavcodec/x86/hevc_mc.asm | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/libavcodec/x86/hevc_mc.asm b/libavcodec/x86/hevc_mc.asm
index 0bb0e17..8444c41 100644
--- a/libavcodec/x86/hevc_mc.asm
+++ b/libavcodec/x86
On 08/01/15 08:27 AM, Henrik Gramner wrote:
The .text section is already 16-byte aligned by default on all supported
platforms so `SECTION_TEXT` isn't any different from `SECTION .text`.
Works fine with -f aout here.
Dave
___
libav-devel mailing list
>From 7aa9c4c5035f777c9c08b49c3af256edf36f4961 Mon Sep 17 00:00:00 2001
From: Dave Yeo
Date: Thu, 16 Jul 2015 21:13:49 -0700
Subject: [PATCH] OS/2:Cleanup slib_create_def_cmd
Export symbols by name rather then ordinal. Remove PROTMODE directive as it
doesn't make sense for 32 bit libra
On 05/10/15 11:05 AM, wm4 wrote:
+#if CONFIG_OPENSSL && HAVE_POLL_H && HAVE_PTHREADS
> >+#include
> >+#include
> >+#include
> >+#define BLOCK_SIGPIPE \
>
>Why an ugly macro rather than two function calls?
It requires at least remembering the previous signal state (the oldset
variable). Since
On 11/23/14 12:31 PM, Anton Khirnov wrote:
Also add no-op fallbacks when threading is disabled.
This helps keeping the code clean if Libav is compiled for targets
without threading. Since we assume that no threads of any kind are used
in such configurations, doing nothing is ok by definition.
B
On 09/23/14 04:22 PM, Diego Biurrun wrote:
On Tue, Sep 23, 2014 at 11:27:28PM +0100, Vittorio Giovara wrote:
On Sun, Sep 21, 2014 at 12:43 PM, Andrew Kelley wrote:
more things to remove?
is there any reason to keep lavf/gopher.c besides compatibility with a
time machine?
Heh, at VDD we had
ection and therefore assembling with NASM
Perhaps revert as attached?
Dave
From 45fc4375a333f4e93a664883a1186e7f6faca9e0 Mon Sep 17 00:00:00 2001
From: Dave Yeo
Date: Sat, 22 Feb 2014 15:44:26 -0800
Subject: [PATCH] configure: revert commit 2b0bb69997c2416e74f41aa1400ce983bf8775c0
This fixes cpunop det
On 11/24/13 01:27 pm, Martin Storsjö wrote:
This fixes building on windows.
---
Fixes OS/2 as well.
What is the plan for --disable-doc? configure --help still states that
it is an option but the build fails with unknown option --disable-doc.
Dave
__
On 10/07/13 09:06 am, Derek Buitenhuis wrote:
From: Henrik Gramner
The Mach-O bug was fixed in yasm 0.8.0 and we don't
support versions that old anymore.
Signed-off-by: Derek Buitenhuis
---
libavutil/x86/x86inc.asm | 16 +---
1 file changed, 5 insertions(+), 11 deletions(-)
Loo
On 09/12/13 03:05 am, Derek Buitenhuis wrote:
On 9/12/2013 2:35 AM, Dave Yeo wrote:
OS/2 still uses a.out from GCC (then converting to OMF32) so we need
a.out support until OMF is added to YASM or our LIBC maintainer finishes
adding ELF support (perhaps halfway done but he doesn't spend
On 09/11/13 08:49 am, Hendrik Leppkes wrote:
From: Henrik Gramner
The Mach-O bug was fixed in yasm 0.8.0 and we don't support versions that old.
a.out was superseded by ELF on sane systems a few decades ago.
OS/2 still uses a.out from GCC (then converting to OMF32) so we need
a.out support u
ype
Possible solution attached
Dave
From 998efd31389f929e0268932446ef0fe08c300afe Mon Sep 17 00:00:00 2001
From: Dave Yeo
Date: Wed, 14 Aug 2013 15:22:13 -0700
Subject: [PATCH] Check for sockaddr_in6 rather then AF_INET6 as some systems such as OS/2 define AF_INET6 without a full implementation
---
libavformat
On 03/16/13 11:59 pm, Reinhard Tartler wrote:
The suppression file has been originally proposed by Reimar Doeffinger.
The idea of using set_defaults is from Luca Barbato.
---
configure|8 ++--
doc/developer.texi |8
tests/fate-valgrind.supp | 31 ++
On 03/02/13 08:57 am, Diego Biurrun wrote:
On Fri, Mar 01, 2013 at 01:38:33PM -0800, Dave Yeo wrote:
On 03/01/13 09:56 am, Luca Barbato wrote:
On 01/03/13 18:53, Vittorio Giovara wrote:
Resend without the unnecessary change.
I'm wondering if we couldn't just drop the support for
On 03/01/13 09:56 am, Luca Barbato wrote:
On 01/03/13 18:53, Vittorio Giovara wrote:
Resend without the unnecessary change.
Vittorio
I'm wondering if we couldn't just drop the support for older sdl and
just rely on pkg-config.
Nothing against the patch but I will point out that not all SDL
On 12/18/12 05:02 pm, Ronald S. Bultje wrote:
From: "Ronald S. Bultje"
---
libavutil/x86/x86inc.asm | 33 -
1 file changed, 20 insertions(+), 13 deletions(-)
Fixes 32 bit NASM assembly here, thanks
Dave
___
libav-de
>From c678894a9cb8ca2dfad1b65709762a12c2033bc2 Mon Sep 17 00:00:00 2001
From: Dave Yeo
Date: Tue, 30 Oct 2012 23:48:26 -0700
Subject: [PATCH] Fix assembly with NASM
After commit 04581c8c77ce779e4e70684ac45302972766be0f all include files need
complete source path
---
libavutil/x86/x86util.
On 10/13/12 10:23 am, Martin Storsjö wrote:
This requires the makedef perl script by Derek, which probably will
go into the c99-to-c89 repo. That scripts produces a .def file,
listing the symbols to be exported, based on the gcc version
scripts and the built object files.
Where is the makedef s
On 10/02/12 07:07 pm, Luca Barbato wrote:
+/**
+ * Symbol visibility support
+ */
+
+#if defined _WIN32 || defined __CYGWIN__
+# define AV_LIBRARY_IMPORT __declspec(dllimport)
+# define AV_LIBRARY_EXPORT __declspec(dllexport)
+#else
+# if __GNUC__>= 4
+# define AV_LIBRARY_IMPORT __att
On 09/04/12 01:47 pm, Martin Storsjö wrote:
+elif check_cpp_condition stddef.h "defined __KLIBC__"; then
+libc_type=klibc
While the patch works fine here, perhaps it should be noted that this is
the Linux klibc and not the OS/2 kLIBC which also has __KLIBC__ defined
though not in any inst
a.out object format does not allow aligning sections. On OS/2 LD aligns
sections to 16 bytes.
---
libavcodec/x86/pngdsp.asm |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/libavcodec/x86/pngdsp.asm b/libavcodec/x86/pngdsp.asm
index d6e6374..970f3db 100644
--- a/libavcode
On 08/06/12 10:47 am, Luca Barbato wrote:
On 8/6/12 7:43 PM, Dave Yeo wrote:
Yasm doesn't support any of the OS/2 object formats (a.out and OMF).
I'm wondering how hard would be adding a.out or omf support to yasm.
There's been a bug open for 7 years about adding omf support
On 08/06/12 09:10 am, Diego Biurrun wrote:
On Mon, Aug 06, 2012 at 03:54:10PM +0100, Måns Rullgård wrote:
Måns Rullgård writes:
Now that's debian. What about other systems, like cygwin? Or AncientBSD?
Cygwin comes with nasm 2.08 (of course you can compile yasm there).
You can compile yas
On 07/08/12 08:26 am, Diego Elio Pettenò wrote:
Il 08/07/2012 17:19, Diego Biurrun ha scritto:
On some systems -lm is not required to link against libmath.
LGTM.
As a side note, IIRC Haiku has (or had) the same need.
OS/2 solves the issue by installing an empty libm.
Dave
_
On 06/26/12 04:57 am, Martin Storsjö wrote:
From: "Ronald S. Bultje"
It is included for the open/read/write/close functions. On
MSVC, where this header does not exist, the same functions
are provided by io.h, which is already included.
If the HAVE_SETMODE for including io.h feels too vague, one
On 06/19/12 11:27 pm, Sean McGovern wrote:
On Wed, Jun 20, 2012 at 2:13 AM, Sean McGovern wrote:
On Wed, Jun 20, 2012 at 2:11 AM, Luca Barbato wrote:
On 06/19/2012 10:04 PM, Sean McGovern wrote:
I'm totally baffled by this response...
You apparently reported the issue on the wrong project
On 06/19/12 02:13 pm, Mans Rullgard wrote:
This builds and passes tests on my linux system. However, some
systems (e.g. BSD) are known to sometimes require explicit inclusion
of headers beyond what the standard mentions, so it would be great
if someone could test this on some such systems.
Sur
On 05/22/12 10:12 pm, Luca Barbato wrote:
On 23/05/12 13:34, Dave Yeo wrote:
---
libavformat/os_support.h |6 ++
1 files changed, 6 insertions(+), 0 deletions(-)
diff --git a/libavformat/os_support.h b/libavformat/os_support.h
index 1088c6c..223203a 100644
--- a/libavformat
n somebody shine a light on this one?
The original error message that my patch addresses is
x86inc.asm:100: error: unknown 'cpu' type
That error is still there with nasm 2.10. Reimar fixed it like with the
attached.
Dave
From e1efa1e5dc69ce2c5808feb47301c73bd69bb2bb Mon Sep 17 00:0
---
libavformat/os_support.h |6 ++
1 files changed, 6 insertions(+), 0 deletions(-)
diff --git a/libavformat/os_support.h b/libavformat/os_support.h
index 1088c6c..223203a 100644
--- a/libavformat/os_support.h
+++ b/libavformat/os_support.h
@@ -45,6 +45,12 @@ static inline int is_dos_pat
On 03/31/12 11:54 am, Ronald S. Bultje wrote:
> tests/ref/fate/bmp-4bit-os2 |2 ++
> tests/ref/fate/bmp-8bit |2 ++
> tests/ref/fate/bmp-8bit-os2 |2 ++
> tests/ref/fate/bmp-rle4 |2 ++
> tests/ref/fate/bmp-rle8 |2 ++
Do want some more OS/2 bmps besides th
On 03/20/12 03:54 am, Mans Rullgard wrote:
> This fixes doc generation in build tree separate from source.
>
> Signed-off-by: Mans Rullgard --- doc/Makefile |
> 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/doc/Makefile b/doc/Makefile index 5172707..a051f6e
> 100644 --
On 12/25/11 03:28 am, Luca Barbato wrote:
On 25/12/11 05:20, Dave Yeo wrote:
While this patch works, the problem is the configure part needs to
#define DOS_INCL. I've added it to the CFLAGS which causes a warning
in pthread.c about DOS_INCL being redefined. I could just check for
os2.
o actually check for the function.
http://cyberkinetica.homeunix.net/os2tk45/addendum/049_L2_DosQuerySysInfo.html
Dave
---
>From a0416f1ed0b428e930ad5c219fd8b1aefbae9a78 Mon Sep 17 00:00:00 2001
From: Dave Yeo
Date: Sat, 24 Dec 2011 20:10:11 -0800
Subject: [PATCH] Detect number of proce
On 11/12/11 10:57 pm, Reinhard Tartler wrote:
Upstream Status (cf.http://www.nasm.us)
The latest stable version of NASM is 2.09.10 (revision history).
The latest release candidate of NASM is 2.10rc8 (revision history).
Note that I'm not sure if the issue has been fixed in 2.10rc8 at all,
On 11/12/11 09:12 pm, Luca Barbato wrote:
On 11/13/11 5:54 AM, Dave Yeo wrote:
From b77189d75915cf3a665f150eaf0f05a617d18ef5 Mon Sep 17 00:00:00 2001
From: Reimar Döffinger
Date: Sat, 12 Nov 2011 20:43:52 -0800
Subject: [PATCH] This issue is actually already fixed in NASM, however
it will be a
ork-around.
Signed-off-by: Reimar Döffinger
Signed-off-by: Dave Yeo
---
libavutil/x86/x86inc.asm | 15 +++
1 files changed, 15 insertions(+), 0 deletions(-)
diff --git a/libavutil/x86/x86inc.asm b/libavutil/x86/x86inc.asm
index 31f74b9..951bdfa 100644
--- a/libavutil/x86/x86inc.asm
++
On 10/24/11 07:20 am, Ronald S. Bultje wrote:
Hi Dave,
On Sun, Oct 23, 2011 at 11:49 PM, Dave Yeo wrote:
On 10/23/11 09:34 am, Ronald S. Bultje wrote:
This fixes assembling using "nasm".
Actually just fixes some warnings.
I use nasm 2.09 for testing. ;-).
I misunderstood
On 10/23/11 09:34 am, Ronald S. Bultje wrote:
This fixes assembling using "nasm".
Actually just fixes some warnings.
"nasm-2.10rc8-20110830-os2" is choking on line 515 of
libavutil/x86/x86inc.asm,
%define RESET_MM_PERMUTATION INIT_MMX
when called from line 429 of scale.asm, so obviously a bug
Hi, with NASM the %include directive can only include files in the
current directory. Fix is to use the -i option to search the correct
directories.
Dave
ps patch should probably be tested on a system with YASM.
diff --git a/Makefile b/Makefile
index 21639ca..28ca055 100644
--- a/Makefile
+++ b/
On 05/28/11 11:39 am, Justin Ruggles wrote:
this should fix the windows builds
---
Could someone test this in mingw32 and/or cygwin to see if this fixes the build?
libavcodec/x86/ac3dsp.asm |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
FYI it fixed the OS/2 build which had
On 05/27/11 11:55 am, Diego Biurrun wrote:
On Thu, May 26, 2011 at 11:08:16PM -0700, Dave Yeo wrote:
On 05/26/11 05:53 am, Diego Biurrun wrote:
Libav should now compile out-of-the-box on OS/2 and the
information contained in the linked page is badly outdated.
It compiles out of the box with
On 05/26/11 05:53 am, Diego Biurrun wrote:
Libav should now compile out-of-the-box on OS/2 and the
information contained in the linked page is badly outdated.
It compiles out of the box with the correct shell ((pdk)sh). Unluckily
eCS ships with ash which has a bug where $PWD is not initially s
On 05/25/11 12:29 pm, Diego Biurrun wrote:
On Thu, May 12, 2011 at 07:02:01PM -0700, Dave Yeo wrote:
On 05/12/11 08:30 am, Diego Biurrun wrote:
Now on to specific FATE configurations and their flags and hacks:
[...]
* os/2
This configuration sets '-D_GNU_SOURCE', which is suspi
This fixes the error,
libavcodec/x86/dct32_sse.asm:193: error: segment name `.text align=16'
not recognized
on OS/2
Dave
ps apologies for not using format-patch, it does not seem to work with
the port of git I'm using.
diff --git a/libavcodec/x86/dct32_sse.asm b/libavcodec/x86/dct32_sse.asm
ind
On 05/20/11 10:10 am, Måns Rullgård wrote:
"Dave Yeo " writes:
diff --git a/configure b/configure
old mode 100755
new mode 100644
What the fuck did you do that for?
Seems to have been a tool chain screw up, only programs compiled against
libc understand *nix permissions whic
On 05/16/11 03:42 am, Luca Barbato wrote:
On 5/15/11 7:42 PM, Dave Yeo wrote:
You need to update your nasm binary as it assembles fine here with the
latest daily and appears to have been fixed in March.
(http://permalink.gmane.org/gmane.comp.lang.nasm.cvs/2465)
The patch seems sensible and
On 05/15/11 06:20 am, Diego Biurrun wrote:
So we test for and appear to support nasm, but nasm fails to build
libavcodec/x86/fft_mmx.asm:
$ make -k V=1
nasm -f win32 -DPREFIX -I. -I"/home/diego/src/libav" -Pconfig.asm -I
libavcodec/x86/ -M -o libavcodec/x86/fft_mmx.o libavcodec/x86/fft_mmx.asm
On 05/12/11 08:30 am, Diego Biurrun wrote:
Now that I have finally fixed the BSD CPPFLAGS issues in a clean way
I believe it is time to review our FATE configurations again and
possibly unify them.
[...]
There are some FATE configurations that use --disable-debug and some
that use --enable-shar
52 matches
Mail list logo
