December 2015 - MPlayer-dev-eng - lists.mplayerhq.hu

Relicensing MPlayer or parts of it to LGPL
by wm4 27 Sep '16

27 Sep '16

I'm maintainer of mpv, a MPlayer fork. One of the changes in mpv was turning the slave mode interface into a library (libmpv), which replaces the stdin/stdout parsing slave mode MPlayer uses. One problem with this is that it forces every client to be GPL. libmpv basically exhibits the slave mode interface as API, but since the application code actually links to it, and the player runs in the same address space, the GPL technically applies (in my interpretation). To avoid this, I'm thinking of relicensing mpv to LGPL 2.1 (with the "or later"). While many things have changed in mpv compared to MPlayer (and many parts have been rewritten or removed), much of its core is still based on MPlayer, and I need agreement from the original authors of the code. I'm thinking it would be beneficial for MPlayer as well to relicense parts to LGPL (or all of it, if enough contributors can be reached). Alternatively, I'd just ask for permission to relicense the affected source files in mpv. Would there be any welcoming or opposing to such a move from MPlayer developers?

8 20

[PATCH] Added support for arm64
by jpinto 24 May '16

24 May '16

Signed-off-by: Joao Pinto <jpinto(a)synopsys.com> Tested-by: Joao Pinto <jpinto(a)synopsys.com> --- configure | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/configure b/configure index 722b8a4..d17e61e 100755 --- a/configure +++ b/configure @@ -2493,6 +2493,11 @@ EOF arch='generic' ;; + aarch64) + arch='arm64' + iproc='arm64' + ;; + *) echo "The architecture of your CPU ($host_arch) is not supported by this configure script" echo "It seems nobody has ported MPlayer to your OS or CPU type yet." -- 1.8.1.5

3 3

Re: [MPlayer-dev-eng] [MPlayer-users] Build regression on FreeBSD
by Alexander Strasser 14 Feb '16

14 Feb '16

Hi Thomas, AFAICT there are multiple problems at work, that make a quick solution rather involved or hackish. That's why I have Cc'ed MPlayer development mailing list. On 2015-12-19 20:21 +0100, Thomas Zander wrote: > porting an up-to-date snapshot to FreeBSD I noticed two build > regressions and want to kindly ask for resolving those upstream: > > 1) Makefile (at root dir) > codec-cfg.c now includes <vdpau/vdpau.h> > On FreeBSD this file ends up in LOCALBASE (by default > /usr/local/include) which is not included in the default compiler > search path, so in its build target > codec-cfg$(EXESUF) .... the compile command > $(HOST_CC) $(HOSTCFLAGS) -o $@ $< > is not sufficient. The include path for localbase needs to be added. I guess r37454 introduced this problem, by including vdpau.h in our img_format.h header. It cannot be reverted because that would probably break the build with future FFmpeg versions. The problem is if we want to use external stuff, which cannot be forward declared, in img_format.h we need to add additional flags to HOSTCFLAGS as well. Furthermore it is interesing that vdpau gets detected at all on your system. I suspect this is because of manual addition to CFLAGS or because of modification by previous configure tests. I might be missing a few things, but all in all I do not see why we need the structure vdpau_frame_data in codec-cfg.c source. So maybe the structure should be defined in a different header? All in all I see no really easy way to set HOSTCFLAGS correctly. Additionally I am not sure if the work is justified, because if I am not mistaken above it is not needed in codec-cfg. [...] Alexander

4 12

Building mplayer with ASAN support
by Gustavo Grieco 07 Feb '16

07 Feb '16

Hello, I'm trying to build the last cve snapshot (export 2015-12-18) with address sanitizer in order to test in deep the last version of mplayer. I used: $ ./configure --extra-cflags="-fsanitize=address" but unfortunately, this is not working: $ make cc -MMD -MP -Wundef -Wall -Wno-switch -Wno-parentheses -Wpointer-arith -Wredundant-decls -Werror=format-security -Wstrict-prototypes -Wmissing-prototypes -Wdisabled-optimization -Wno-pointer-sign -Wdeclaration-after-statement -std=gnu99 -Werror-implicit-function-declaration -D_POSIX_C_SOURCE=200112 -D_XOPEN_SOURCE=600 -D_ISOC99_SOURCE -I. -Iffmpeg -O4 -march=native -mtune=native -pipe -ffast-math -fomit-frame-pointer -fno-tree-vectorize -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_LARGEFILE64_SOURCE -fsanitize=address -fpie -DPIC -D_REENTRANT -I/usr/include/p11-kit-1 -D_REENTRANT -I/usr/include/freetype2 -DZLIB_CONST -c -o command.o command.c In file included from command.c:27:0: config.h:124:23: warning: missing terminating " character [enabled by default] #define EXTERN_PREFIX "__asan_init_v1 U ^ config.h:125:1: warning: missing terminating " character [enabled by default] " ^ config.h:125:1: error: missing terminating " character In file included from stream/stream.h:22:0, from command.c:30: ./config.h:124:23: warning: missing terminating " character [enabled by default] #define EXTERN_PREFIX "__asan_init_v1 U ^z ./config.h:125:1: warning: missing terminating " character [enabled by default] " ^ In file included from ./mp_msg.h:143:0, from stream/stream.h:24, from command.c:30: ./config.h:124:23: warning: missing terminating " character [enabled by default] #define EXTERN_PREFIX "__asan_init_v1 U ^ ./config.h:125:1: warning: missing terminating " character [enabled by default] " ^ command.c: In function 'mp_property_sub': command.c:1453:22: warning: variable 'reset_spu' set but not used [-Wunused-but-set-variable] int source = -1, reset_spu = 0; ^ make: *** [command.o] Error 1 If i disable the custom CFLAGS, mplayer will compile correctly. Is ASAN supported somehow? Thanks!

2 1

MPlayer 1.2.1 release
by Roberto Togni 27 Jan '16

27 Jan '16

Hi I plan to release 1.2.1 in a few days. 1.2.1 will be made from the 1.2 branch, and will include FFmpeg 2.8.4 In addition to what is already backported, I'd like to include also these changes, even if they are not really bug fixes - r37545 compile with external FFmpeg without requiring x86/asm.h This should make life easier for distributions - r37549 creation of docs without requiring FFmpeg Also useful for distributions, already included in Debian - r37546 silence warning from FFmpeg about network initialization I'm undecided about this, but should be harmless Opinions? If you have anything else that you think is relevant for inclusion in 1.2.1 please commit it, or replay this message. Ciao, Roberto

5 11

How to report a couple of crashes with malformed wav files
by Gustavo Grieco 25 Jan '16

25 Jan '16

Hello, I'm the main developer and maintainer of QuickFuzz, a free and open-source experimental grammar fuzzer. I recently made a quick test of Mplayer version shipped with Ubuntu 14.04 and found a few interesting crashes trying to play malformed wav files. These crashes are de-duplicated by Honggfuzz, so they should be more or less independent (although, some are definitely related). The list is here: SIGBUS.PC.5704be.STACK.5c9a551.CODE.128.ADDR.(nil).INSTR.callq__0xfffffffffff3c6e2.fuzz SIGFPE.PC.432d0e.STACK.18b3c0fcd4.CODE.1.ADDR.0x432d0e.INSTR.idivl__0x560ca8(%rip)________#_0x0000000000560cae.fuzz SIGFPE.PC.4bf2c3.STACK.1bca543b66.CODE.1.ADDR.0x4bf2c3.INSTR.divl___0x80(%rsi).fuzz SIGFPE.PC.7ffff3ceed83.STACK.d7d8808dd.CODE.1.ADDR.0x7ffff3ceed83.INSTR.idiv___%r8d.fuzz SIGSEGV.PC.4bd833.STACK.18b2dd10ac.CODE.1.ADDR.0xa.INSTR.movzwl_0xa(%rax),%ecx.fuzz SIGSEGV.PC.4be46b.STACK.1aba63653d.CODE.1.ADDR.(nil).INSTR.movzbl_(%rcx,%rdx,1),%ecx.fuzz SIGSEGV.PC.5704be.STACK.5c9a551.CODE.1.ADDR.0x7ffffaaf0e08.INSTR.callq__0xfffffffffff3c6e2.fuzz SIGSEGV.PC.5704be.STACK.5c9a551.CODE.1.ADDR.0x7ffffefcecd8.INSTR.callq__0xfffffffffff3c6e2.fuzz SIGSEGV.PC.5704be.STACK.5c9a551.CODE.1.ADDR.0x7fffff48eff8.INSTR.callq__0xfffffffffff3c6e2.fuzz SIGSEGV.PC.5704be.STACK.5c9a551.CODE.1.ADDR.0x7fffff54e258.INSTR.callq__0xfffffffffff3c6e2.fuzz SIGSEGV.PC.5704be.STACK.5c9a551.CODE.1.ADDR.0x7fffff60c128.INSTR.callq__0xfffffffffff3c6e2.fuzz SIGSEGV.PC.5704be.STACK.5c9a551.CODE.1.ADDR.0x7fffff60d498.INSTR.callq__0xfffffffffff3c6e2.fuzz SIGSEGV.PC.5704be.STACK.5c9a551.CODE.1.ADDR.0x7fffff60d4a8.INSTR.callq__0xfffffffffff3c6e2.fuzz SIGSEGV.PC.5704be.STACK.5c9a551.CODE.1.ADDR.0x7fffff60daa8.INSTR.callq__0xfffffffffff3c6e2.fuzz SIGSEGV.PC.5704be.STACK.5c9a551.CODE.1.ADDR.0x7fffff60db58.INSTR.callq__0xfffffffffff3c6e2.fuzz SIGSEGV.PC.5704be.STACK.5c9a551.CODE.1.ADDR.0x7fffff6429a8.INSTR.callq__0xfffffffffff3c6e2.fuzz SIGSEGV.PC.5947b5.STACK.1b0df30f87.CODE.1.ADDR.(nil).INSTR.movzbl_(%rcx,%rdx,1),%ecx.fuzz SIGSEGV.PC.7fffefe7314e.STACK.18f3cc3594.CODE.1.ADDR.(nil).INSTR.movdqu_%xmm8,(%rdi,%rcx,1).fuzz I think some of them can be security issues (that's why they are not linked in this email). I want to handle such crashes to some trusted developers of Mplayer (i found no email address or contact form for security issues) and avoid spamming the bug tracker. Which is the best way to collaborate with the Mplayer project? Regards, Gus.

3 7

[PATCH] Fix attribute parsing at SRT subtitles
by Stefan Pöschel 18 Jan '16

18 Jan '16

Hi, when SRT subtitles shall be displayed including formatting, they are converted to ASS markup. The conversion therefore also processes the "color" attribute of the HTML "font" tag. At the moment unfortunately also affects attributes who just have "color" as a prefix, e.g. the attribute "back-color" which is sometimes contained in subtitles. Therefore the value of this background color attribute is by mistake used to set the text color. Please find attached a test SRT file (correct text color: white, not blue). The attached patch fixes this parsing by also including the leading space when recognizing the "color" attribute and also the other attributes. Thanks and regards, Stefan Pöschel

2 2

[PATCH] gui/win32: Port X11 GUI changes of r37501
by Alexander Strasser 23 Dec '15

23 Dec '15

Support dynamic label variable $O (title name or file name). Commit-r37501-by: Ingo Brückl --- I fear guiInfo.Title is never set in win32 GUI, so that this does not really add new functionality. Still it seems better to have this patch in, as currently the label variable $O is not evaluated at all, IOW practically we have a loss of functionality, which was the initial reason why I started looking into this. gui/win32/widgetrender.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) --- a/gui/win32/widgetrender.c +++ b/gui/win32/widgetrender.c @@ -128,7 +128,8 @@ static void stringreplace(char *dest, const char *what, const char *format, ... * * @param how 0 (cut file path and extension), * 1 (additionally, convert lower case) or - * 2 (additionally, convert upper case) + * 2 (additionally, convert upper case) or + * 4 (unaltered title if available, otherwise like 0) * @param fname memory location of a buffer to receive the converted Filename * @param maxlen size of the @a fname buffer * @@ -144,7 +145,9 @@ static char *TranslateFilename (int how, char *fname, size_t maxlen) { case STREAMTYPE_FILE: - if (guiInfo.Filename && *guiInfo.Filename) + if ((how == 4) && guiInfo.Title) + av_strlcpy(fname, guiInfo.Title, maxlen); + else if (guiInfo.Filename && *guiInfo.Filename) { p = strrchr(guiInfo.Filename, '\\'); @@ -237,6 +240,7 @@ static char *generatetextfromlabel(widget *item) stringreplace(text, "$D", "%3.0f", guiInfo.Balance); stringreplace(text, "$t", "%.2i", guiInfo.Track); stringreplace(text, "$o", "%s", acp(TranslateFilename(0, tmp, sizeof(tmp)))); + stringreplace(text, "$O", "%s", acp(TranslateFilename(4, tmp, sizeof(tmp)))); stringreplace(text, "$x", "%i", guiInfo.VideoWidth); stringreplace(text, "$y", "%i", guiInfo.VideoHeight); stringreplace(text, "$C", "%s", guiInfo.sh_video ? codecname : ""); --

2 1

Re: [MPlayer-dev-eng] [MPlayer-users] Build regression on FreeBSD
by Ingo Brückl 23 Dec '15

23 Dec '15

I wrote on Sat, 19 Dec 2015 20:50:18 +0100: > Thomas Zander wrote on Sat, 19 Dec 2015 20:21:08 +0100: >> 2) stream/stream_vcd.c: >> when building with libcdio, vcd_read_fbsd.h is no longer included, >> which, in turn, causes for sys/cdrio.h not to be included. >> Subsequently CDRIOCSETBLOCKSIZE is an undefined symbol and >> compilation breaks. Including sys/cdrio.h in stream/stream_vcd.c >> resolves the problem and the build completes. > I suppose we don't need the ioctl CDRIOCSETBLOCKSIZE in case of libcdio, > do we? When libcdio is used, it should take care and the call should be > unnecessary. I found the ioctl CDRIOCSETBLOCKSIZE call in libcdio so we should go with the attached patch. Ingo

1 0

[PATCH 0/3] mingw/msys improvements
by Alexander Strasser 21 Dec '15

21 Dec '15

Hi all, this is what I came out when I was building MPlayer with mingw64 on msys2. Some additional comments follow in the summary below. Alexander Strasser (2): configure: iconv: Remove hack that added -L/usr/lib This pulled in cygwin/msys2 libs and ruined the build. The layout is /mingw64/lib for mingw64 and /usr/lib and msys stuff. Does anyone still need/use that hack? configure: Detect SChannel and enable FFmpeg https SChannel autodetection in configure modelled after GnuTLS check. 1) I do not know if CONFIG_GNUTLS was needed in the first place, need to check what FFmpeg configure generates. It was added first in SVN r34326 . 2) I am not sure why I added pkg-config flags in the enable case too. It's different from how almost all other configure checks behave. Reimar Döffinger (1): configure: Drastically speed up generation of config.h/.asm/.mak Thank you Reimar for looking into this. On my machines it saves about half a minute on Linux and about 10 minutes on Windows. The generated files are somewhat messier because they always include "define XXX 0" for all items. I am not sure we should care much about that ATM, but it can be kind of confusing when searching/grepping those files. configure | 45 +++++++++++++++++++++++++++++++-------------- 1 file changed, 31 insertions(+), 14 deletions(-) -- Alexander

2 11