headers
Loren Merritt | 2 Jun 2008 17:39

Re: [PATCH] x264 executable stack and selinux

On Mon, 12 May 2008, Guillaume POIRIER wrote:
>
> However, I'm tempted to reject that patch since for some time already,
> only YASM support has been supported, so having binaries built with
> NASM that cause some trouble is to be expected.
>
> Maybe a cleared statement or a configure test should be added to
> require YASM, but I doubt that it will please people, since IIRC,
> simple cases such as static lib on Linux and Windows x86 (that is, a
> large majority of x264's userbase) should work, thought it's mainly by
> luck.
>
> I'd prefer if Loren could take a decision about the problem of NASM's support.

nasm works for me.

--Loren Merritt
_______________________________________________
x264-devel mailing list
x264-devel <at> videolan.org
http://mailman.videolan.org/listinfo/x264-devel
Permalink | Reply | 1 comment |
headers
Guillaume POIRIER | 2 Jun 2008 17:53
Picon
Gravatar

Re: [PATCH] x264 executable stack and selinux

Hello,

On 6/2/08, Loren Merritt <lorenm <at> u.washington.edu> wrote:
> On Mon, 12 May 2008, Guillaume POIRIER wrote:
>  >
>  > However, I'm tempted to reject that patch since for some time already,
>  > only YASM support has been supported, so having binaries built with
>  > NASM that cause some trouble is to be expected.
>  >
>  > Maybe a cleared statement or a configure test should be added to
>  > require YASM, but I doubt that it will please people, since IIRC,
>  > simple cases such as static lib on Linux and Windows x86 (that is, a
>  > large majority of x264's userbase) should work, thought it's mainly by
>  > luck.
>  >
>  > I'd prefer if Loren could take a decision about the problem of NASM's support.
>
>
> nasm works for me.

Then I must have made some confusion in my mind with platform-specific
limitations/bug, e.g. on OSX 10.4, yasm is required to build the
dynamic version of x264's lib (static version works fine).

So what do you say Loren, do you like that patch?

Guillaume
Permalink | Reply | 0 comments |
headers
git version control | 2 Jun 2008 20:15
Favicon

commit: Fix noise reduction in threaded mode. Previously enabling noise reduction with threads had no effect. ( Alexander Strange )

x264 | branch: master | Alexander Strange <astrange <at> ithinksw.com> | Mon Jun  2 12:16:51 2008 -0600| [f7e1d82a90b0ec21e1fda5842cf5146bdb6125dc]

Fix noise reduction in threaded mode.  Previously enabling noise reduction with threads had no effect.

> http://git.videolan.org/gitweb.cgi/x264.git/?a=commit;h=f7e1d82a90b0ec21e1fda5842cf5146bdb6125dc
---

 encoder/encoder.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/encoder/encoder.c b/encoder/encoder.c
index 39c5d5f..aa9d372 100644
--- a/encoder/encoder.c
+++ b/encoder/encoder.c
 <at>  <at>  -1602,7 +1602,7  <at>  <at>  static void x264_encoder_frame_end( x264_t *h, x264_t *thread_current,
     /* restore CPU state (before using float again) */
     x264_emms();

-    x264_noise_reduction_update( h );
+    x264_noise_reduction_update( thread_current );

     /* ---------------------- Compute/Print statistics --------------------- */
     x264_thread_sync_stat( h, h->thread[0] );

_______________________________________________
x264-devel mailing list
x264-devel <at> videolan.org
http://mailman.videolan.org/listinfo/x264-devel
(Continue reading)

Permalink | Reply | 0 comments |
headers
git version control | 2 Jun 2008 20:38
Favicon

commit: Fix noise reduction in threaded mode. (Alexander Strange )

x264 | branch: master | Alexander Strange <astrange <at> ithinksw.com> | Mon Jun  2 12:16:51 2008 -0600| [c415d98820c96b8eb489736b1c9ed7c5927b077f]

Fix noise reduction in threaded mode.
Previously enabling noise reduction with threads had no effect.
Note that this is not an optimal solution; each thread still tracks noise reducation separately (unlike in
single-threaded mode).

> http://git.videolan.org/gitweb.cgi/x264.git/?a=commit;h=c415d98820c96b8eb489736b1c9ed7c5927b077f
---

 encoder/encoder.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/encoder/encoder.c b/encoder/encoder.c
index 39c5d5f..aa9d372 100644
--- a/encoder/encoder.c
+++ b/encoder/encoder.c
 <at>  <at>  -1602,7 +1602,7  <at>  <at>  static void x264_encoder_frame_end( x264_t *h, x264_t *thread_current,
     /* restore CPU state (before using float again) */
     x264_emms();

-    x264_noise_reduction_update( h );
+    x264_noise_reduction_update( thread_current );

     /* ---------------------- Compute/Print statistics --------------------- */
     x264_thread_sync_stat( h, h->thread[0] );

_______________________________________________
x264-devel mailing list
x264-devel <at> videolan.org
(Continue reading)

Permalink | Reply | 1 comment |
headers
git version control | 2 Jun 2008 20:53
Favicon

commit: 2-pass VBV support and improved VBV handling (Gabriel Bouvigne )

x264 | branch: master | Gabriel Bouvigne <gabriel.bouvigne <at> joost.com> | Mon Jun  2 12:53:01 2008 -0600| [7331ea4e381898421a1a908ed57f73126faacd36]

2-pass VBV support and improved VBV handling
Dramatically improves 1-pass VBV ratecontrol (especially CBR) and provides support for VBV in 2-pass
mode.  This consists of a series of functions that attempts to find overflows and underflows in the VBV from
the first-pass statsfile and fix them before encoding.
1-pass VBV code partially by Dark Shikari.

> http://git.videolan.org/gitweb.cgi/x264.git/?a=commit;h=7331ea4e381898421a1a908ed57f73126faacd36
---

 common/frame.c        |   28 ++++++
 common/frame.h        |    3 +
 encoder/encoder.c     |    1 +
 encoder/ratecontrol.c |  239 ++++++++++++++++++++++++++++++++++++++++++++-----
 encoder/ratecontrol.h |    2 +
 5 files changed, 250 insertions(+), 23 deletions(-)

Diff:   http://git.videolan.org/gitweb.cgi/x264.git/?a=commitdiff;h=7331ea4e381898421a1a908ed57f73126faacd36
_______________________________________________
x264-devel mailing list
x264-devel <at> videolan.org
http://mailman.videolan.org/listinfo/x264-devel
Permalink | Reply | 0 comments |
headers
Gabriel Bouvigne | 3 Jun 2008 09:15
Favicon

Re: [PATCH] x264 executable stack and selinux

Paul -Spawn- Rysevets a écrit :

> GP> Then I must have made some confusion in my mind with platform-specific
> GP> limitations/bug, e.g. on OSX 10.4, yasm is required to build the
> GP> dynamic version of x264's lib (static version works fine).
> 
> and under Vista x64 Enterprise under MSVS 2008 ... latest nightly of
> nasm crashes on cabac-a.asm and sometimes goes to infinity loop with
> deblock-a.asm

There is definitively an issue regarding deblocking with curent versions 
compiled under MSVC, NASM or YASM is not the issue by itself.

--
Gabriel

_______________________________________________
x264-devel mailing list
x264-devel <at> videolan.org
http://mailman.videolan.org/listinfo/x264-devel
Permalink | Reply | 0 comments |
headers
git version control | 4 Jun 2008 00:53
Favicon

commit: fix some uninitialized partitions in rdo (Noboru Asai )

x264 | branch: master | Noboru Asai <noboru.asai <at> gmail.com> | Mon Jun  2 09:12:29 2008 -0600| [26088fc4526d329a12296b862f10c85fb7801d2f]

fix some uninitialized partitions in rdo

> http://git.videolan.org/gitweb.cgi/x264.git/?a=commit;h=26088fc4526d329a12296b862f10c85fb7801d2f
---

 encoder/analyse.c |    6 ++++++
 1 files changed, 6 insertions(+), 0 deletions(-)

diff --git a/encoder/analyse.c b/encoder/analyse.c
index dad24b1..c0c44b3 100644
--- a/encoder/analyse.c
+++ b/encoder/analyse.c
 <at>  <at>  -1874,6 +1874,7  <at>  <at>  static void x264_mb_analyse_p_rd( x264_t *h, x264_mb_analysis_t *a, int i_satd )
     if( a->l0.i_cost8x8 <= thresh )
     {
         h->mb.i_type = P_8x8;
+        h->mb.i_partition = D_8x8;
         x264_analyse_update_cache( h, a );
         a->l0.i_cost8x8 = x264_rd_cost_mb( h, a->i_lambda2 );

 <at>  <at>  -1892,6 +1893,7  <at>  <at>  static void x264_mb_analyse_p_rd( x264_t *h, x264_mb_analysis_t *a, int i_satd )
             {
                 h->mb.i_sub_partition[0] = h->mb.i_sub_partition[1] =
                 h->mb.i_sub_partition[2] = h->mb.i_sub_partition[3] = D_L0_8x8;
+                x264_analyse_update_cache( h, a );
                 i_cost = x264_rd_cost_mb( h, a->i_lambda2 );
                 if( a->l0.i_cost8x8 < i_cost )
                 {
(Continue reading)

Permalink | Reply | 0 comments |
headers
git version control | 4 Jun 2008 00:53
Favicon

commit: allow fractional CRF values with AQ. (Jason Garrett-Glaser )

x264 | branch: master | Jason Garrett-Glaser <darkshikari <at> gmail.com> | Wed Apr  9 16:46:51 2008 -0600| [9a9d39a568cf7a9486cc04014e22a4fc99fc0b7f]

allow fractional CRF values with AQ.

> http://git.videolan.org/gitweb.cgi/x264.git/?a=commit;h=9a9d39a568cf7a9486cc04014e22a4fc99fc0b7f
---

 encoder/ratecontrol.c |    8 ++++++--
 1 files changed, 6 insertions(+), 2 deletions(-)

diff --git a/encoder/ratecontrol.c b/encoder/ratecontrol.c
index 030d0a0..e071726 100644
--- a/encoder/ratecontrol.c
+++ b/encoder/ratecontrol.c
 <at>  <at>  -77,6 +77,7  <at>  <at>  struct x264_ratecontrol_t
     ratecontrol_entry_t *rce;
     int qp;                     /* qp for current frame */
     int qpm;                    /* qp for current macroblock */
+    float f_qpm;                /* qp for current macroblock: precise float for AQ */
     float qpa_rc;               /* average of macroblocks' qp before aq */
     float qpa_aq;               /* average of macroblocks' qp after aq */
     int qp_force;
 <at>  <at>  -245,9 +246,9  <at>  <at>  void x264_autosense_aq( x264_t *h )
 *****************************************************************************/
 void x264_adaptive_quant( x264_t *h )
 {
-    int qp = h->mb.i_qp;
     int energy = ac_energy_mb( h, h->mb.i_mb_x, h->mb.i_mb_y, NULL );
     /* Adjust the QP based on the AC energy of the macroblock. */
+    float qp = h->rc->f_qpm;
(Continue reading)

Permalink | Reply | 0 comments |
headers
git version control | 4 Jun 2008 00:53
Favicon

commit: omit redundant mc after non-rdo dct size decision, and in b-direct rdo (Jason Garrett-Glaser )

x264 | branch: master | Jason Garrett-Glaser <darkshikari <at> gmail.com> | Fri May 23 21:22:29 2008 -0600| [f88d09450264d597e7b08f83993bed95ba1f7cfa]

omit redundant mc after non-rdo dct size decision, and in b-direct rdo

> http://git.videolan.org/gitweb.cgi/x264.git/?a=commit;h=f88d09450264d597e7b08f83993bed95ba1f7cfa
---

 common/common.h      |    2 +-
 common/macroblock.c  |    2 +-
 encoder/analyse.c    |   24 +++++++++++++++++-------
 encoder/macroblock.c |   11 ++++++-----
 encoder/me.c         |    2 +-
 5 files changed, 26 insertions(+), 15 deletions(-)

diff --git a/common/common.h b/common/common.h
index 425fc8f..bb72957 100644
--- a/common/common.h
+++ b/common/common.h
 <at>  <at>  -429,7 +429,7  <at>  <at>  struct x264_t
         int i_skip_intra;
         /* skip flag for P/B-skip motion compensation */
         /* if we've already done skip MC, we don't need to do it again */
-        int b_skip_pbskip_mc;
+        int b_skip_mc;

         struct
         {
diff --git a/common/macroblock.c b/common/macroblock.c
index 7c8b939..d640664 100644
--- a/common/macroblock.c
(Continue reading)

Permalink | Reply | 0 comments |
headers
git version control | 4 Jun 2008 00:53
Favicon

commit: binmode stdin on mingw, not just msvc (BugMaster )

x264 | branch: master | BugMaster <BugMaster <at> narod.ru> | Sat May 24 01:09:07 2008 -0600| [02a04f2d03e80be3f66bb76fd1bc775216b0d37c]

binmode stdin on mingw, not just msvc

> http://git.videolan.org/gitweb.cgi/x264.git/?a=commit;h=02a04f2d03e80be3f66bb76fd1bc775216b0d37c
---

 common/osdep.h |    5 ++++-
 x264.c         |    2 +-
 2 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/common/osdep.h b/common/osdep.h
index c7353e0..18c7c48 100644
--- a/common/osdep.h
+++ b/common/osdep.h
 <at>  <at>  -31,9 +31,12  <at>  <at> 
 #include <inttypes.h>
 #endif

-#ifdef _MSC_VER
+#ifdef _WIN32
 #include <io.h>    // _setmode()
 #include <fcntl.h> // _O_BINARY
+#endif
+
+#ifdef _MSC_VER
 #define inline __inline
 #define strcasecmp stricmp
 #define strncasecmp strnicmp
diff --git a/x264.c b/x264.c
(Continue reading)

Permalink | Reply | 0 comments |

Gmane