From gb.public at free.fr Thu Oct 2 21:35:03 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Thu, 2 Oct 2008 23:35:03 +0200 Subject: [PATCH] Fix for Flash 10 windowless mode In-Reply-To: <48E25318.9000901@redhat.com> References: <9570662.1356521222759094777.JavaMail.root@spooler5-g27.priv.proxad.net> <48E25318.9000901@redhat.com> Message-ID: <064369B3-AE50-4B7A-81FD-9ADDCFA4AC8F@free.fr> Hi, Le 30 sept. 08 ? 18:26, Warren Togami a ?crit : > http://www.theonion.com/content/video/economists_warn_anti_bush > This video seems to crash readily if a Flash ad happens in rotation > while the video is playing. The bug can't be reproduced reliably but I know what's the cause: NPN_InvalidateRect() is called from an idle or timer function but the wrapper could already be sending whatever NPP_*() function at the same time. The latter receives an MSG_START for NPN_InvalidateRect() whereas it was expecting an MSG_ACK for the NPP_*() function it sent. Here is a scenario to make things clear: the browser calls NPP_Write() -> rpc_method_invoke(): - send: MSG_START NPP_WRITE MSG_END - wait: MSG_ACK However, the viewer was executing a timer function that calls NPN_Invalidate() - send: MSG_START NPN_INVALIDATE_RECT MSG_END - wait: MSG_ACK and the browser receives that MSG_START. Two possible solutions. 1) The easiest but not the most efficient one: queue NPN_InvalidateRect() and send them on a browser request. e.g. browser sets a timer function that regularily calls into the viewer to get pending messages. In this model, the viewer shall never decides itself to send messages if it was not initiated from the browser. 2) (Re)introduce a two-channels communication model. It was problematic in the initial implementation but I think I can get it right this time. We will maintain this dual rule: - NPN communication channel always calls into the viewer (NPP_*() function calls). This is a bidirectional channel: viewer can call into the browser provided the communication was initiated by the browser. e.g. NPN_UserAgent() will be handled there. - NPP communication channel always calls into the browser (NPN_*() function calls). However, the range of functions are limited to those that the viewer can initiate from a timer for example. In practise, those are NPN_InvalidateRect() and NPW_UpdateDrawable(). The NPP communication channel would be run in a separate thread in the viewer. The one in the browser would be hooked into a GSource. I will look into solution 2. Regards, Gwenol?. From wtogami at redhat.com Fri Oct 3 17:19:20 2008 From: wtogami at redhat.com (Warren Togami) Date: Fri, 03 Oct 2008 13:19:20 -0400 Subject: Youtube stops during playback Message-ID: <48E65418.6000409@redhat.com> flash-plugin-10.0.0.569-release.i386 firefox-3.0.2-1.fc9.x86_64 nspluginwrapper-1.1.0-7.fc9.i386 nspluginwrapper-1.1.0-7.fc9.x86_64 [warren at newcaprica ~]$ cat /etc/adobe/mms.cfg WindowlessDisable=true After that latest patch for windowless mode, I noticed that Youtube and other Flash videos play and seem to randomly stop mid-way. ../src/npw-viewer.c:2735../src/npw-viewer.c:773../src/npw-wrapper.c:294../src/npw-wrapper.c:286../src/npw-wrapper.c:288../src/npw-viewer.c:708../src/npw-viewer.c:728../src/npw-viewer.c:734../src/npw-viewer.c:2737../src/npw-wrapper.c:1491../src/npw-wrapper.c:1489../src/npw-viewer.c:2747../src/npw-viewer.c:2735 stdout from the browser is printing things like this a great many times in a loop during video before it stops. Downgraded to nspluginwrapper-1.1.0-5.fc9 without your patch and video playing works without this failure. Also the rapid printing of the above stdout does not happen during video play. Warren Togami wtogami at redhat.com From wtogami at redhat.com Fri Oct 3 22:27:48 2008 From: wtogami at redhat.com (Warren Togami) Date: Fri, 03 Oct 2008 18:27:48 -0400 Subject: Youtube stops during playback In-Reply-To: <48E65418.6000409@redhat.com> References: <48E65418.6000409@redhat.com> Message-ID: <48E69C64.50803@redhat.com> Warren Togami wrote: > Downgraded to nspluginwrapper-1.1.0-5.fc9 without your patch and video > playing works without this failure. Also the rapid printing of the > above stdout does not happen during video play. > It seems that nspluginwrapper-1.1.0-7.fc9 never reached mirrors, so I am not building a reverted package for F9. I am buliding -10.fc10 for rawhide though. Warren Togami wtogami at redhat.com From wtogami at redhat.com Sun Oct 5 15:08:23 2008 From: wtogami at redhat.com (Warren Togami) Date: Sun, 05 Oct 2008 11:08:23 -0400 Subject: Youtube stops during playback In-Reply-To: <48E65418.6000409@redhat.com> References: <48E65418.6000409@redhat.com> Message-ID: <48E8D867.2040109@redhat.com> Warren Togami wrote: > > Downgraded to nspluginwrapper-1.1.0-5.fc9 without your patch and video > playing works without this failure. Also the rapid printing of the > above stdout does not happen during video play. Sigh. Now the same Youtube failure is happening with nspluginwrapper-1.1.0-5.fc9. I don't know what is going on. Perhaps this latest patch for -7 is actually not to blame. /src/npw-wrapper.c:1904../src/npw-viewer.c:3053../src/npw-viewer.c:3055../src/npw-wrapper.c:1906../src/npw-wrapper.c:1771../src/npw-viewer.c:2980../src/npw-viewer.c:2962../src/npw-viewer.c:2964../src/npw-wrapper.c:1773../src/npw-wrapper.c:1665../src/npw-wrapper.c:1667 -7 is incredibly noisy with stuff like this. It seems the video randomly stopping issue can happen with -5 or -7 if pulseaudio had somehow crashed. So I was wrong. Why is it so noisy? Warren From gb.public at free.fr Sun Oct 5 22:09:03 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Mon, 6 Oct 2008 00:09:03 +0200 Subject: [PATCH] Fix for concurrent rpc_method_invoke() In-Reply-To: <064369B3-AE50-4B7A-81FD-9ADDCFA4AC8F@free.fr> References: <9570662.1356521222759094777.JavaMail.root@spooler5-g27.priv.proxad.net> <48E25318.9000901@redhat.com> <064369B3-AE50-4B7A-81FD-9ADDCFA4AC8F@free.fr> Message-ID: <3B01CEFF-F152-4C57-950D-755C47924DC2@free.fr> Hi, Le 2 oct. 08 ? 23:35, Gwenole Beauchesne a ?crit : > I will look into solution 2. Finally, I changed my mind and went for solution 3. ;-) Reminder of the problem: - browser calls rpc_method_invoke() and waits for an MSG_ACK prior to sending the arguments - plugin can call rpc_method_invoke() at the same time Thus, instead of the expected MSG_ACK, the browser gets a new MSG_START message. This strategy was useful as an optimization, both run-time and writing- time (yes, the human using the API ;-)). It turned out to be a broken approach as it permitted an unexpected foreign call from the plugin. I (wrongly?) assumed that any NPN_*() function call from the plugin shall be made from NPN context, i.e. "we get to call into an NPN_*() function only if the plugin is executing code within an NPP_*() function called by the browser". In practise, this is no longer true with Flash 10 since the following functions can be called from a timer: NPN_GetURLNotify(), NPN_UserAgent(), NPN_InvalidateRect(). Now, rpc_method_invoke() writes the method ID and its arguments in one shot. This means the above-mentioned optimization vanishes. Though, it might be reintroduced in V3 RPC protocol specification for a future nspluginwrapper version. Here is the patch I committed to SVN, along with testcases. It applies on top of the previous patch I posted (to fix windowless mode). Note that I have only tested with Firefox 3.1-snapshot and the latest Flash 10 RC. More testing is welcome. Thanks. Regards, Gwenole. -------------- next part -------------- A non-text attachment was scrubbed... Name: npw.rpc.v2.5.diff Type: application/octet-stream Size: 49381 bytes Desc: not available URL: -------------- next part -------------- From wtogami at redhat.com Sun Oct 5 23:44:08 2008 From: wtogami at redhat.com (Warren Togami) Date: Sun, 05 Oct 2008 19:44:08 -0400 Subject: [PATCH] Fix for concurrent rpc_method_invoke() In-Reply-To: <3B01CEFF-F152-4C57-950D-755C47924DC2@free.fr> References: <9570662.1356521222759094777.JavaMail.root@spooler5-g27.priv.proxad.net> <48E25318.9000901@redhat.com> <064369B3-AE50-4B7A-81FD-9ADDCFA4AC8F@free.fr> <3B01CEFF-F152-4C57-950D-755C47924DC2@free.fr> Message-ID: <48E95148.2000904@redhat.com> Hunk #1 succeeded at 137 (offset -29 lines). Hunk #2 succeeded at 187 (offset -3 lines). patch unexpectedly ends in middle of line patch: **** malformed patch at line 1724: Is the patch you sent malformed? Warren From gb.public at free.fr Mon Oct 6 04:45:41 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Mon, 6 Oct 2008 06:45:41 +0200 Subject: [PATCH] Fix for concurrent rpc_method_invoke() In-Reply-To: <48E95148.2000904@redhat.com> References: <9570662.1356521222759094777.JavaMail.root@spooler5-g27.priv.proxad.net> <48E25318.9000901@redhat.com> <064369B3-AE50-4B7A-81FD-9ADDCFA4AC8F@free.fr> <3B01CEFF-F152-4C57-950D-755C47924DC2@free.fr> <48E95148.2000904@redhat.com> Message-ID: <50AA82D5-8287-41EF-9174-3CC6D2FB5D0A@free.fr> Le 6 oct. 08 ? 01:44, Warren Togami a ?crit : > Hunk #1 succeeded at 137 (offset -29 lines). > Hunk #2 succeeded at 187 (offset -3 lines). > patch unexpectedly ends in middle of line > patch: **** malformed patch at line 1724: > > Is the patch you sent malformed? It was manually edited to remove excess from svn diff -r646:HEAD but it applies fine here (on top of rev 646). Hmm, I edited it on MacOS X so it probably contains extra UTF-8 characters? Here is the same one with a ChangeLog entry but regenerated from within Linux. I don't know if that'd really help though. -------------- next part -------------- A non-text attachment was scrubbed... Name: npw.rpc.v2.5.diff Type: application/octet-stream Size: 51118 bytes Desc: not available URL: -------------- next part -------------- From gb.public at free.fr Mon Oct 6 07:07:01 2008 From: gb.public at free.fr (gb.public at free.fr) Date: Mon, 6 Oct 2008 09:07:01 +0200 (CEST) Subject: [PATCH] Fix for concurrent rpc_method_invoke() In-Reply-To: <50AA82D5-8287-41EF-9174-3CC6D2FB5D0A@free.fr> Message-ID: <24421638.1867011223276821288.JavaMail.root@spooler5-g27.priv.proxad.net> Hi, ----- "Gwenole Beauchesne" a ?crit : > Here is the same one with a ChangeLog entry but regenerated from > within Linux. I don't know if that'd really help though. FYI, I could apply it on top of your package from: though I have not tried to compile it. Regards, Gwenole. From asac at jwsdot.com Mon Oct 6 07:44:56 2008 From: asac at jwsdot.com (Alexander Sack) Date: Mon, 6 Oct 2008 09:44:56 +0200 Subject: [PATCH] Fix for concurrent rpc_method_invoke() In-Reply-To: <50AA82D5-8287-41EF-9174-3CC6D2FB5D0A@free.fr> References: <9570662.1356521222759094777.JavaMail.root@spooler5-g27.priv.proxad.net> <48E25318.9000901@redhat.com> <064369B3-AE50-4B7A-81FD-9ADDCFA4AC8F@free.fr> <3B01CEFF-F152-4C57-950D-755C47924DC2@free.fr> <48E95148.2000904@redhat.com> <50AA82D5-8287-41EF-9174-3CC6D2FB5D0A@free.fr> Message-ID: <20081006074456.GA21005@jwsdot.com> Hi, On Mon, Oct 06, 2008 at 06:45:41AM +0200, Gwenole Beauchesne wrote: > Le 6 oct. 08 ? 01:44, Warren Togami a ?crit : > >> Hunk #1 succeeded at 137 (offset -29 lines). >> Hunk #2 succeeded at 187 (offset -3 lines). >> patch unexpectedly ends in middle of line >> patch: **** malformed patch at line 1724: >> >> Is the patch you sent malformed? > > It was manually edited to remove excess from svn diff -r646:HEAD but it > applies fine here (on top of rev 646). Hmm, I edited it on MacOS X so it > probably contains extra UTF-8 characters? > > Here is the same one with a ChangeLog entry but regenerated from within > Linux. I don't know if that'd really help though. this patch applied more or less cleanly to our ubuntu packages. I think it works better now. Here an excerpt of from the console output i still see. Are those VisualID errors because I do not have the latest patches? BTW, did you publish your SVN somewhere? .... (firefox:27970): GLib-CRITICAL **: g_hash_table_insert_internal: assertion `hash_table != NULL' failed *** NSPlugin Wrapper *** ERROR: NPP_NewStream() invoke: Connection closed (firefox:27970): GLib-CRITICAL **: g_hash_table_insert_internal: assertion `hash_table != NULL' failed *** NSPlugin Wrapper *** ERROR: NPP_NewStream() invoke: Connection closed NPP_GetValue() NPP_GetValue() NPP_GetValue() NPP_GetValue() *** NSPlugin Wrapper *** ERROR: NPP_SetWindow() invoke: Connection closed *** NSPlugin Wrapper *** ERROR: NPP_HandleEvent() invoke: Connection closed *** NSPlugin Wrapper *** ERROR: NPP_SetWindow() invoke: Connection closed *** NSPlugin Wrapper *** ERROR: NPP_HandleEvent() invoke: Connection closed *** NSPlugin Wrapper *** ERROR: NPP_HandleEvent() invoke: Connection closed *** NSPlugin Wrapper *** ERROR: NPP_HandleEvent() invoke: Connection closed NPP_GetValue() NPP_GetValue() NPP_GetValue() NPP_GetValue() *** NSPlugin Wrapper *** ERROR: NPP_HandleEvent() invoke: Connection closed *** NSPlugin Wrapper *** ERROR: NPP_HandleEvent() invoke: Connection closed *** NSPlugin Wrapper *** ERROR: NPP_SetWindow() invoke: Connection closed *** NSPlugin Wrapper *** ERROR: NPP_HandleEvent() invoke: Connection closed *** NSPlugin Viewer *** ERROR: could not reconstruct XVisual from visualID *** NSPlugin Viewer *** ERROR: could not reconstruct XVisual from visualID *** NSPlugin Viewer *** ERROR: could not reconstruct XVisual from visualID *** NSPlugin Viewer *** ERROR: could not reconstruct XVisual from visualID *** NSPlugin Viewer *** ERROR: could not reconstruct XVisual from visualID *** NSPlugin Viewer *** ERROR: could not reconstruct XVisual from visualID - Alexander From asac at jwsdot.com Mon Oct 6 07:52:13 2008 From: asac at jwsdot.com (Alexander Sack) Date: Mon, 6 Oct 2008 09:52:13 +0200 Subject: [PATCH] Fix for concurrent rpc_method_invoke() In-Reply-To: <20081006074456.GA21005@jwsdot.com> References: <9570662.1356521222759094777.JavaMail.root@spooler5-g27.priv.proxad.net> <48E25318.9000901@redhat.com> <064369B3-AE50-4B7A-81FD-9ADDCFA4AC8F@free.fr> <3B01CEFF-F152-4C57-950D-755C47924DC2@free.fr> <48E95148.2000904@redhat.com> <50AA82D5-8287-41EF-9174-3CC6D2FB5D0A@free.fr> <20081006074456.GA21005@jwsdot.com> Message-ID: <20081006075213.GB21005@jwsdot.com> On Mon, Oct 06, 2008 at 09:44:56AM +0200, Alexander Sack wrote: > Hi, > > On Mon, Oct 06, 2008 at 06:45:41AM +0200, Gwenole Beauchesne wrote: > > Le 6 oct. 08 ? 01:44, Warren Togami a ?crit : > > > >> Hunk #1 succeeded at 137 (offset -29 lines). > >> Hunk #2 succeeded at 187 (offset -3 lines). > >> patch unexpectedly ends in middle of line > >> patch: **** malformed patch at line 1724: > >> > >> Is the patch you sent malformed? > > > > It was manually edited to remove excess from svn diff -r646:HEAD but it > > applies fine here (on top of rev 646). Hmm, I edited it on MacOS X so it > > probably contains extra UTF-8 characters? > > > > Here is the same one with a ChangeLog entry but regenerated from within > > Linux. I don't know if that'd really help though. > > this patch applied more or less cleanly to our ubuntu packages. I > think it works better now. > > Here an excerpt of from the console output i still see. Are those > VisualID errors because I do not have the latest patches? OK, I added the * src/npw-viewer.c (create_window_attributes): Get system visual when we are passed a 0 visual ID. hunk posted earlier and the visualID issues are gone. Works even better. Thanks. Any idea where the "Message type invalid" might stem from? *** NSPlugin Viewer *** ERROR: NPN_GetValue() invoke: Message type invalid Also the following output looks a bit scary? *** NSPlugin Wrapper *** ERROR: NPObject 0x44ac7b0 is not valid! Here the complete console output: *** NSPlugin Viewer *** ERROR: NPN_GetValue() invoke: Message type invalid *** NSPlugin Viewer *** WARNING: unhandled variable 14 in NPN_GetValue() *** NSPlugin Viewer *** ERROR: NPN_InvalidateRect() invoke: Message type invalid *** NSPlugin Viewer *** ERROR: NPN_InvalidateRect() invoke: Message type invalid *** NSPlugin Viewer *** ERROR: NPN_InvalidateRect() invoke: Message type invalid *** NSPlugin Wrapper *** ERROR: passing an unknown instance rpc_method_invoke: Connection reset by peer *** NSPlugin Wrapper *** ERROR: NPClass::Invalidate() invoke: Connection reset by peer *** NSPlugin Wrapper *** ERROR: NPP_Destroy() invoke: Connection closed (firefox:30934): GLib-CRITICAL **: g_hash_table_remove_internal: assertion `hash_table != NULL' failed *** NSPlugin Wrapper *** ERROR: NPObject 0x44ac7b0 is not valid! *** NSPlugin Wrapper *** ERROR: NPP_Destroy() invoke: Connection closed (firefox:30934): GLib-CRITICAL **: g_hash_table_remove_internal: assertion `hash_table != NULL' failed *** NSPlugin Wrapper *** ERROR: NPObject 0x24551e0 is not valid! *** NSPlugin Wrapper *** ERROR: NPP_Destroy() invoke: Connection closed (firefox:30934): GLib-CRITICAL **: g_hash_table_remove_internal: assertion `hash_table != NULL' failed *** NSPlugin Wrapper *** ERROR: NPObject 0x3f8ff10 is not valid! *** NSPlugin Wrapper *** ERROR: NPP_Destroy() invoke: Connection closed - Alexander From gb.public at free.fr Mon Oct 6 08:51:24 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Mon, 6 Oct 2008 10:51:24 +0200 (CEST) Subject: [PATCH] Fix for concurrent rpc_method_invoke() In-Reply-To: <16646208.1882281223282937176.JavaMail.root@spooler5-g27.priv.proxad.net> Message-ID: <27854902.1882661223283084037.JavaMail.root@spooler5-g27.priv.proxad.net> Hi, ----- "Alexander Sack" a ?crit : > hunk posted earlier and the visualID issues are gone. Works even > better. Thanks. Any idea where the "Message type invalid" might stem > from? Well, the latest patch was supposed to fix this kind of errors... :-/ Are you sure you have re-wrapped the plugins? I forgot to mention since RPC code changed, both the wrapper and the viewer must be in sync. I hope this is the error you are facing. I had 7 tabs playing plenty of Flash content without problem (and wow, a Core 2 Duo @ 2 GHz suffers). BTW, the SVN was available here for some time: Or as ViewVC: Regards, Gwenole. From wtogami at redhat.com Mon Oct 6 21:34:34 2008 From: wtogami at redhat.com (Warren Togami) Date: Mon, 06 Oct 2008 17:34:34 -0400 Subject: Test Results: Windowless + Concurrent rpc_method_invoke() patch Message-ID: <48EA846A.1090403@redhat.com> nspluginwrapper-1.1.0 plus Gwenole's fixes for windowless and the Concurrent rpc_method_invoke() patch test results. http://www.theonion.com/content/video/obama_promises_to_stop_americas The Onion videos no longer seem to crash in limited testing. http://news.google.com/ Youtube videos embedded in Google News no longer seem to crash if windowless mode is enabled. However I did notice an intermittent problem. Sometimes right-click over a Flash video fails to pop-up Flash's context menu. Overall, it is an improvement with these two patches. Gwenole, why am I seeing a tremendous amount of debug error printing being dumped to stdout (or is that stderr?) during video playback? Anyhow, I am building this for Fedora 10 as nspluginwrapper-1.1.0-11.fc10 for rawhide only for now. I hope that we can fix whatever is causing the tremendous amount of stuff being printed before we push a Fedora 9 update. Probably a good time to release nspluginwrapper-1.1.1 after that is fixed? Warren Togami wtogami at redhat.com From wtogami at redhat.com Mon Oct 6 22:20:37 2008 From: wtogami at redhat.com (Warren Togami) Date: Mon, 06 Oct 2008 18:20:37 -0400 Subject: Test Results: Windowless + Concurrent rpc_method_invoke() patch In-Reply-To: <48EA846A.1090403@redhat.com> References: <48EA846A.1090403@redhat.com> Message-ID: <48EA8F35.9090205@redhat.com> Warren Togami wrote: > http://news.google.com/ > Youtube videos embedded in Google News no longer seem to crash if > windowless mode is enabled. However I did notice an intermittent > problem. Sometimes right-click over a Flash video fails to pop-up > Flash's context menu. > OK, this right-click problem is actually 100% reproducible. If windowless mode is enabled (default in Flash 10) then the embedded Youtube video on Google News exhibits this problem. If you disable it with /etc/adobe/mms.cfg then the context menu works. It works unwrapped. Warren Togami wtogami at redhat.com From ktp420 at live.com Mon Oct 6 22:28:46 2008 From: ktp420 at live.com (Ketan Patel) Date: Mon, 6 Oct 2008 18:28:46 -0400 Subject: Another testcase which always crashes npviewer for... Message-ID: Hi, I have reported this on following ubuntu bug: https://bugs.launchpad.net/ubuntu/hardy/+source/ia32-libs/+bug/192888/comments/322 Another testcase in which npviewer.bin stops when starting to play another clip.... goto dialbo III website and click on to see different skill and once one is done playing the next one should start but npviewer.bin process stops/crashes/quites...no npviewer.bin process and no flash on the page. http://www.blizzard.com/diablo3/characters/witchdoctor.xml Can someone confirm the latest patches fixes the issues on diablo3 site? I have also noticed most of the time flash stop when I scroll up and down in the browser. Hope it helps. Ketan _________________________________________________________________ See how Windows connects the people, information, and fun that are part of your life. http://clk.atdmt.com/MRT/go/msnnkwxp1020093175mrt/direct/01/ From wtogami at redhat.com Mon Oct 6 23:51:24 2008 From: wtogami at redhat.com (Warren Togami) Date: Mon, 06 Oct 2008 19:51:24 -0400 Subject: Another testcase which always crashes npviewer for... In-Reply-To: References: Message-ID: <48EAA47C.8070902@redhat.com> Ketan Patel wrote: > Hi, > > I have reported this on following ubuntu bug: > https://bugs.launchpad.net/ubuntu/hardy/+source/ia32-libs/+bug/192888/comments/322 > > Another testcase in which npviewer.bin stops when starting to play another clip.... > > goto dialbo III website and click on to see different skill and once one is done playing the next one should start but npviewer.bin process stops/crashes/quites...no npviewer.bin process and no flash on the page. > > http://www.blizzard.com/diablo3/characters/witchdoctor.xml > > Can someone confirm the latest patches fixes the issues on diablo3 site? I have also noticed most of the time flash stop when I scroll up and down in the browser. > I just watched through all 5 skill videos in a row and it seems working here. firefox-3.0.2-1.fc9.x86_64 flash-plugin-10.0.0.569-release.i386 nspluginwrapper-1.1.0-11.fc9.i386 nspluginwrapper-1.1.0-11.fc9.x86_64 Warren Togami wtogami at redhat.com From ktp420 at live.com Tue Oct 7 02:45:07 2008 From: ktp420 at live.com (Ketan Patel) Date: Mon, 6 Oct 2008 22:45:07 -0400 Subject: Another testcase which always crashes npviewer for... In-Reply-To: <48EAA47C.8070902@redhat.com> References: <48EAA47C.8070902@redhat.com> Message-ID: Thanks Warren, Alexander let me know if you have a deb with the patch you want me to test. I wouldn't mind taking test drive. Thanks Gwenole for the patch. ---------------------------------------- > Date: Mon, 6 Oct 2008 19:51:24 -0400 > From: wtogami at redhat.com > To: nspluginwrapper-devel-list at redhat.com > Subject: Re: Another testcase which always crashes npviewer for... > > Ketan Patel wrote: >> Hi, >> >> I have reported this on following ubuntu bug: >> https://bugs.launchpad.net/ubuntu/hardy/+source/ia32-libs/+bug/192888/comments/322 >> >> Another testcase in which npviewer.bin stops when starting to play another clip.... >> >> goto dialbo III website and click on to see different skill and once one is done playing the next one should start but npviewer.bin process stops/crashes/quites...no npviewer.bin process and no flash on the page. >> >> http://www.blizzard.com/diablo3/characters/witchdoctor.xml >> >> Can someone confirm the latest patches fixes the issues on diablo3 site? I have also noticed most of the time flash stop when I scroll up and down in the browser. >> > > I just watched through all 5 skill videos in a row and it seems working > here. > > firefox-3.0.2-1.fc9.x86_64 > flash-plugin-10.0.0.569-release.i386 > nspluginwrapper-1.1.0-11.fc9.i386 > nspluginwrapper-1.1.0-11.fc9.x86_64 > > Warren Togami > wtogami at redhat.com > > _______________________________________________ > Nspluginwrapper-devel-list mailing list > Nspluginwrapper-devel-list at redhat.com > https://www.redhat.com/mailman/listinfo/nspluginwrapper-devel-list _________________________________________________________________ See how Windows connects the people, information, and fun that are part of your life. http://clk.atdmt.com/MRT/go/msnnkwxp1020093175mrt/direct/01/ From gb.public at free.fr Tue Oct 7 05:33:23 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Tue, 7 Oct 2008 07:33:23 +0200 Subject: Test Results: Windowless + Concurrent rpc_method_invoke() patch In-Reply-To: <48EA846A.1090403@redhat.com> References: <48EA846A.1090403@redhat.com> Message-ID: <9799009F-65A6-45B1-9CD3-2F71843D288D@free.fr> Hi, Le 6 oct. 08 ? 23:34, Warren Togami a ?crit : > Gwenole, why am I seeing a tremendous amount of debug error printing > being dumped to stdout (or is that stderr?) during video playback? I don't know yet. I looked at my sources and could not notice any debug code with __FILE__:__LINE__ in npw-viewer.c. It's probably coming from one of your patches? I will look at yours for what it matches tonight. Have you tried to run with NPW_DEBU=1, any more info? Since you are missing part of the message, it's probably just debug code that is using npw_printf() instead of npw_dprintf() for example? > Anyhow, I am building this for Fedora 10 as > nspluginwrapper-1.1.0-11.fc10 for rawhide only for now. I hope that > we can fix whatever is causing the tremendous amount of stuff being > printed before we push a Fedora 9 update. Probably a good time to > release nspluginwrapper-1.1.1 after that is fixed? I would like to fix the power saving patch too. Actually, the patch you are using is causing two regressions (two plugins). I have fixed one, I have not fixed the other yet. I will also add the remaining patches from Transitive and it should be fine for a release. Regards, Gwenole. From gb.public at free.fr Tue Oct 7 19:36:35 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Tue, 7 Oct 2008 21:36:35 +0200 (CEST) Subject: [PATCH] Re: Youtube stops during playback In-Reply-To: <3876293.2063721223408159378.JavaMail.root@spooler5-g27.priv.proxad.net> Message-ID: <9651329.2063831223408195985.JavaMail.root@spooler5-g27.priv.proxad.net> Hi, ----- "Warren Togami" a ?crit : > Why is it so noisy? Debug code in your tree got into the released package. Here is a patch to your nspluginwrapper-1.1.0-build.patch. Basically, just remove the extra fprintf(stderr,...) and it should be fine. ;-) Regards, Gwenol?. -------------- next part -------------- A non-text attachment was scrubbed... Name: npw.redhat.debug.diff Type: text/x-patch Size: 1213 bytes Desc: not available URL: From gb.public at free.fr Tue Oct 7 20:44:23 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Tue, 7 Oct 2008 22:44:23 +0200 Subject: Test Results: Windowless + Concurrent rpc_method_invoke() patch In-Reply-To: <48EA8F35.9090205@redhat.com> References: <48EA846A.1090403@redhat.com> <48EA8F35.9090205@redhat.com> Message-ID: Hi, Le 7 oct. 08 ? 00:20, Warren Togami a ?crit : > OK, this right-click problem is actually 100% reproducible. If > windowless mode is enabled (default in Flash 10) then the embedded > Youtube video on Google News exhibits this problem. If you disable > it with /etc/adobe/mms.cfg then the context menu works. Yes, I can confirm this only happens with windowless plugins. i.e. if the plugin does not request wmode=transparent, then the context menu is shown. I have not investigated further yet. A fake XEvent probably should be sent on the viewer side after we caught NPP_HandleEvent() so that Gtk signal handlers could be called? Regards, Gwenole. From gb.public at free.fr Tue Oct 7 20:51:32 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Tue, 7 Oct 2008 22:51:32 +0200 Subject: Test Results: Windowless + Concurrent rpc_method_invoke() patch In-Reply-To: References: <48EA846A.1090403@redhat.com> <48EA8F35.9090205@redhat.com> Message-ID: <62708ACD-432C-416D-92B8-AB5956FD3B47@free.fr> Le 7 oct. 08 ? 22:44, Gwenole Beauchesne a ?crit : >> OK, this right-click problem is actually 100% reproducible. If >> windowless mode is enabled (default in Flash 10) then the embedded >> Youtube video on Google News exhibits this problem. If you disable >> it with /etc/adobe/mms.cfg then the context menu works. > > Yes, I can confirm this only happens with windowless plugins. i.e. > if the plugin does not request wmode=transparent, then the context > menu is shown. I have not investigated further yet. A fake XEvent > probably should be sent on the viewer side after we caught > NPP_HandleEvent() so that Gtk signal handlers could be called? Err, on second thought, by definition, a windowless plugin would not have a GtkWindow. From ktp420 at live.com Thu Oct 9 20:42:55 2008 From: ktp420 at live.com (Ketan Patel) Date: Thu, 9 Oct 2008 16:42:55 -0400 Subject: Test Results: Windowless + Concurrent rpc_method_invoke() patch In-Reply-To: <62708ACD-432C-416D-92B8-AB5956FD3B47@free.fr> References: <48EA846A.1090403@redhat.com> <48EA8F35.9090205@redhat.com> <62708ACD-432C-416D-92B8-AB5956FD3B47@free.fr> Message-ID: I have also notice that middle while does not work when over flash area...i.e. use middle while to scroll up and down a page, but when over flash area it will not scroll. ---------------------------------------- > From: gb.public at free.fr > To: nspluginwrapper-devel-list at redhat.com > Date: Tue, 7 Oct 2008 22:51:32 +0200 > Subject: Re: Test Results: Windowless + Concurrent rpc_method_invoke() patch > > Le 7 oct. 08 ? 22:44, Gwenole Beauchesne a ?crit : > >>> OK, this right-click problem is actually 100% reproducible. If >>> windowless mode is enabled (default in Flash 10) then the embedded >>> Youtube video on Google News exhibits this problem. If you disable >>> it with /etc/adobe/mms.cfg then the context menu works. >> >> Yes, I can confirm this only happens with windowless plugins. i.e. >> if the plugin does not request wmode=transparent, then the context >> menu is shown. I have not investigated further yet. A fake XEvent >> probably should be sent on the viewer side after we caught >> NPP_HandleEvent() so that Gtk signal handlers could be called? > > Err, on second thought, by definition, a windowless plugin would not > have a GtkWindow. > > _______________________________________________ > Nspluginwrapper-devel-list mailing list > Nspluginwrapper-devel-list at redhat.com > https://www.redhat.com/mailman/listinfo/nspluginwrapper-devel-list _________________________________________________________________ See how Windows Mobile brings your life together?at home, work, or on the go. http://clk.atdmt.com/MRT/go/msnnkwxp1020093182mrt/direct/01/ From wtogami at redhat.com Sat Oct 11 00:51:07 2008 From: wtogami at redhat.com (Warren Togami) Date: Fri, 10 Oct 2008 20:51:07 -0400 Subject: Test Results: Windowless + Concurrent rpc_method_invoke() patch In-Reply-To: References: <48EA846A.1090403@redhat.com> <48EA8F35.9090205@redhat.com> <62708ACD-432C-416D-92B8-AB5956FD3B47@free.fr> Message-ID: <48EFF87B.2050007@redhat.com> Ketan Patel wrote: > I have also notice that middle while does not work when over flash > area...i.e. use middle while to scroll up and down a page, but when > over flash area it will not scroll. Hasn't that always been true of any browser plugin, even before nspluginwrapper existed? Warren From gb.public at free.fr Sun Oct 12 22:55:13 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Mon, 13 Oct 2008 00:55:13 +0200 Subject: [ANN] nspluginwrapper 1.1.2 Message-ID: <30C6060B-780E-4262-A799-8B55B1B7009F@free.fr> Hi, I have just released nspluginwrapper 1.1.2. Here is a short list of changes. * Add support for Open Solaris hosts * Add support for ARM targets (Geraint North) * Fix support for windowless plugins (Flash Player 10 rc) * Fix various bugs in RPC code (crashes and concurrent messaging) * Allow wrapping of native plugins through the -n|--native option As you can see, I did not include everything I initially planned. After all, those fixes are already important enough to release early and 1.1.x is a development branch anyway. So, there is some food left for the next release. Some further details are available here: Regards, Gwenole. From gb.public at free.fr Sun Oct 12 23:03:19 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Mon, 13 Oct 2008 01:03:19 +0200 Subject: Test Results: Windowless + Concurrent rpc_method_invoke() patch In-Reply-To: <48EFF87B.2050007@redhat.com> References: <48EA846A.1090403@redhat.com> <48EA8F35.9090205@redhat.com> <62708ACD-432C-416D-92B8-AB5956FD3B47@free.fr> <48EFF87B.2050007@redhat.com> Message-ID: <05426A1E-F7D6-4760-AAFB-3120DF8388BB@free.fr> Hi, Le 11 oct. 08 ? 02:51, Warren Togami a ?crit : > Ketan Patel wrote: >> I have also notice that middle while does not work when over flash >> area...i.e. use middle while to scroll up and down a page, but when >> over flash area it will not scroll. > > Hasn't that always been true of any browser plugin, even before > nspluginwrapper existed? I have tried with an upstream 32-bit Firefox build, scrolling does work over a Flash area. So, this is indeed a bug in nspluginwrapper. Ketan, has this ever worked for you? For other plugins too? IOW, I don't think it's a regression and the bug was probably there from the begining. Regards, Gwenole. From ktp420 at live.com Sun Oct 12 23:42:28 2008 From: ktp420 at live.com (Ketan Patel) Date: Sun, 12 Oct 2008 19:42:28 -0400 Subject: Test Results: Windowless + Concurrent rpc_method_invoke() patch In-Reply-To: <05426A1E-F7D6-4760-AAFB-3120DF8388BB@free.fr> References: <48EA846A.1090403@redhat.com> <48EA8F35.9090205@redhat.com> <62708ACD-432C-416D-92B8-AB5956FD3B47@free.fr> <48EFF87B.2050007@redhat.com> <05426A1E-F7D6-4760-AAFB-3120DF8388BB@free.fr> Message-ID: Gwenole, I don't think this is a regression either. The scrolling also does not work with totem browser plugin on my current system. Ketan ---------------------------------------- > From: gb.public at free.fr > To: nspluginwrapper-devel-list at redhat.com > Date: Mon, 13 Oct 2008 01:03:19 +0200 > Subject: Re: Test Results: Windowless + Concurrent rpc_method_invoke() patch > > Hi, > > Le 11 oct. 08 ? 02:51, Warren Togami a ?crit : > >> Ketan Patel wrote: >>> I have also notice that middle while does not work when over flash >>> area...i.e. use middle while to scroll up and down a page, but when >>> over flash area it will not scroll. >> >> Hasn't that always been true of any browser plugin, even before >> nspluginwrapper existed? > > I have tried with an upstream 32-bit Firefox build, scrolling does > work over a Flash area. So, this is indeed a bug in nspluginwrapper. > Ketan, has this ever worked for you? For other plugins too? IOW, I > don't think it's a regression and the bug was probably there from the > begining. > > Regards, > Gwenole. > > _______________________________________________ > Nspluginwrapper-devel-list mailing list > Nspluginwrapper-devel-list at redhat.com > https://www.redhat.com/mailman/listinfo/nspluginwrapper-devel-list _________________________________________________________________ Want to do more with Windows Live? Learn ?10 hidden secrets? from Jamie. http://windowslive.com/connect/post/jamiethomson.spaces.live.com-Blog-cns!550F681DAD532637!5295.entry?ocid=TXT_TAGLM_WL_domore_092008 From wtogami at redhat.com Mon Oct 13 00:18:37 2008 From: wtogami at redhat.com (Warren Togami) Date: Sun, 12 Oct 2008 20:18:37 -0400 Subject: Warren away for ~2.5 weeks Message-ID: <48F293DD.3070103@redhat.com> Soon I will be travelling on business for a solid 2.5 weeks. Martin Stransky (the actual Fedora nspluginwrapper maintainer) will take care of integrating Fedora's patches with the new 1.1.2 release. Martin, why do we have the debug print patch that Gwenole pointed out? Warren From demontager at gmail.com Sat Oct 18 17:19:04 2008 From: demontager at gmail.com (Andrey) Date: Sat, 18 Oct 2008 14:19:04 -0300 Subject: Flash10 Message-ID: <48FA1A88.6040506@gmail.com> Hi! I can't make nspluginwrapper working with flash10. I was triyng to compile wrapper from sources but no success. From Ubuntu repo 0.9 wrapper availabe but when i give him command to install flash10 it says no compatible plugins found. How to solve this in Ubuntu 64? --- Andrey From demontager at gmail.com Sat Oct 18 17:24:39 2008 From: demontager at gmail.com (Andrey) Date: Sat, 18 Oct 2008 14:24:39 -0300 Subject: Flash10 Message-ID: <48FA1BD7.7090107@gmail.com> Hi! I can't make nspluginwrapper working with flash10. I was triyng to compile wrapper from sources but no success. From Ubuntu repo 0.9 wrapper availabe but when i give him command to install flash10 it says no compatible plugins found. How to solve this in Ubuntu 64? --- Andrey From ktp420 at live.com Sun Oct 19 21:32:17 2008 From: ktp420 at live.com (Ketan Patel) Date: Sun, 19 Oct 2008 17:32:17 -0400 Subject: Another testcase that still fails for me with the latest... In-Reply-To: <48FA1BD7.7090107@gmail.com> References: <48FA1BD7.7090107@gmail.com> Message-ID: Hi Gwenole, I am still seeing crashes when trying to use the grandcenral: http://www.grandcentral.com The first time I goto the site, flash navigation works. But if I click on other tabs, i.e Inbox, Address Book..., the new page has gray box where flash should be. If you don't get the gray box, then keep clicking..try going back to already clicked tab. Thanks, Ketan _________________________________________________________________ When your life is on the go?take your life with you. http://clk.atdmt.com/MRT/go/115298558/direct/01/ From wtogami at redhat.com Sun Oct 19 23:15:55 2008 From: wtogami at redhat.com (Warren Togami) Date: Sun, 19 Oct 2008 19:15:55 -0400 Subject: nspluginwrapper-1.1.2 invisible video Message-ID: <48FBBFAB.6070003@redhat.com> flash-plugin-10.0.12.36-release.i386 nspluginwrapper-1.1.2-1.fc9.i386 nspluginwrapper-1.1.2-1.fc9.x86_64 firefox-3.0.2-1.fc9.x86_64 http://www.theonion.com/content/video/john_mccain_accidentally_left_on If you reload a few times (or sometimes the first time), it somehow gets into a state where you can hear the audio playing, but there is a blank white area where the Flash video should be. Right-clicking the area where the video should be shows an outline of the video, but no menus or anything. Reloading the page fails to recover the video view. [root at newcaprica ~]# cat /etc/adobe/mms.cfg WindowlessDisable=true Disabling WMODE in the plugin seems to allow the video to play. Warren Togami wtogami at redhat.com From asac at jwsdot.com Mon Oct 20 09:20:02 2008 From: asac at jwsdot.com (Alexander Sack) Date: Mon, 20 Oct 2008 11:20:02 +0200 Subject: nspluginwrapper-1.1.2 invisible video In-Reply-To: <48FBBFAB.6070003@redhat.com> References: <48FBBFAB.6070003@redhat.com> Message-ID: <20081020092002.GA24654@jwsdot.com> On Sun, Oct 19, 2008 at 07:15:55PM -0400, Warren Togami wrote: > flash-plugin-10.0.12.36-release.i386 > nspluginwrapper-1.1.2-1.fc9.i386 > nspluginwrapper-1.1.2-1.fc9.x86_64 > firefox-3.0.2-1.fc9.x86_64 > > http://www.theonion.com/content/video/john_mccain_accidentally_left_on > If you reload a few times (or sometimes the first time), it somehow gets > into a state where you can hear the audio playing, but there is a blank > white area where the Flash video should be. Right-clicking the area > where the video should be shows an outline of the video, but no menus or > anything. Reloading the page fails to recover the video view. > > [root at newcaprica ~]# cat /etc/adobe/mms.cfg > WindowlessDisable=true > > Disabling WMODE in the plugin seems to allow the video to play. I sometimes see something similar in a different context. However, i dont get sound. Just a grey area. Looking at the process. I ran with NPW_DEBUG=... and it appears that the new wrapper didnt get further than NP_Initialize when reloading the site. Here the last lines showing the NP_Shutdown (plugin_exit) and the failed restart. ... *** NSPlugin Viewer *** NPP_WriteReady instance=0x8892830, stream=0x890eae8 *** NSPlugin Viewer *** return: 268435455 *** NSPlugin Wrapper *** return: 268435455 *** NSPlugin Wrapper *** NPP_Write instance=0xa847750 *** NSPlugin Viewer *** NPP_Write instance=0x8892830, stream=0x890eae8, offset=0, len=1200, buf=0x88fdbb0 *** NSPlugin Viewer *** return: 1200 *** NSPlugin Wrapper *** return: 1200 *** NSPlugin Wrapper *** NPP_DestroyStream instance=0xa847750 *** NSPlugin Viewer *** handle_NPP_DestroyStream *** NSPlugin Viewer *** NPP_DestroyStream instance=0x8892830, stream=0x890eae8, reason=NPRES_DONE *** NSPlugin Viewer *** return: 0 [NPERR_NO_ERROR] *** NSPlugin Wrapper *** return: 0 [NPERR_NO_ERROR] *** NSPlugin Wrapper *** NPP_URLNotify instance=0xa847750, url='http://i3.ytimg.com/vi/RikNFufdrsc/default.jpg', reason=0, notifyData=0x1e *** NSPlugin Wrapper *** done *** NSPlugin Wrapper *** NPP_NewStream instance=0xa847750 *** NSPlugin Viewer *** NPP_NewStream instance=0x8892830, stream=0x88ca118, url='http://i4.ytimg.com/vi/SXyxy2l-vLQ/default.jpg', type='image/jpeg', seekable=0, stype=NP_NORMAL, notifyData=0x1f *** NSPlugin Viewer *** NPN_UserAgent instance=0x8892830 *** NSPlugin Viewer *** user_agent='Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.0.3) Gecko/2008101315 Ubuntu/8.10 (intrepid) Firefox/3.0.3' *** NSPlugin Viewer *** NPN_GetValue instance=0x8892830, variable=15 [0000000f] *** NSPlugin Wrapper *** handle_NPN_GetValue *** NSPlugin Wrapper *** NPN_GetValue instance=0xa847750, variable=15 *** NSPlugin Wrapper *** return: 0 [NPERR_NO_ERROR] *** NSPlugin Viewer *** value: 0x88fdbd8 *** NSPlugin Viewer *** NPN_GetStringIdentifier name='location' *** NSPlugin Viewer *** return: 0x7 *** NSPlugin Viewer *** NPN_GetProperty instance=0x8892830, npobj=0x88fdbd8, propertyName=0x7 *** NSPlugin Viewer *** return: 1 () *** NSPlugin Viewer *** NPN_GetStringIdentifier name='toString' *** NSPlugin Viewer *** return: 0x8 *** NSPlugin Viewer *** NPN_Invoke instance=0x8892830, npobj=0x88fdbf8, methodName=0x8 *** NSPlugin Viewer *** 0 args () *** NSPlugin Viewer *** return: 1 ('http://www.youtube.com/') *** NSPlugin Viewer *** NPN_ReleaseVariantValue *** NSPlugin Viewer *** done *** NSPlugin Viewer *** NPN_ReleaseObject 0x88fdbd8 *** NSPlugin Viewer *** return: 0 *** NSPlugin Viewer *** NPN_ReleaseVariantValue *** NSPlugin Viewer *** NPN_ReleaseObject 0x88fdbf8 *** NSPlugin Viewer *** return: 0 *** NSPlugin Viewer *** done *** NSPlugin Viewer *** return: 0 [NPERR_NO_ERROR], stype=NP_NORMAL *** NSPlugin Viewer *** NPN_InvalidateRect instance=0x8892830 *** NSPlugin Wrapper *** return: 0 [NPERR_NO_ERROR], stype=NP_NORMAL *** NSPlugin Wrapper *** NPP_WriteReady instance=0xa847750 *** NSPlugin Wrapper *** handle_NPN_InvalidateRect *** NSPlugin Wrapper *** NPN_InvalidateRect instance=0xa847750 *** NSPlugin Wrapper *** done *** NSPlugin Viewer *** done *** NSPlugin Viewer *** NPN_InvalidateRect instance=0x8892830 *** NSPlugin Viewer *** handle_NPP_WriteReady *** NSPlugin Viewer *** NPP_WriteReady instance=0x8892830, stream=0x88ca118 *** NSPlugin Viewer *** return: 268435455 *** NSPlugin Wrapper *** return: 268435455 *** NSPlugin Wrapper *** NPP_Write instance=0xa847750 *** NSPlugin Wrapper *** handle_NPN_InvalidateRect *** NSPlugin Wrapper *** NPN_InvalidateRect instance=0xa847750 *** NSPlugin Wrapper *** done *** NSPlugin Viewer *** done *** NSPlugin Viewer *** NPN_InvalidateRect instance=0x8892830 *** NSPlugin Viewer *** NPP_Write instance=0x8892830, stream=0x88ca118, offset=0, len=1338, buf=0x88fdbd8 *** NSPlugin Viewer *** return: 1338 *** NSPlugin Wrapper *** return: 1338 *** NSPlugin Wrapper *** NPP_DestroyStream instance=0xa847750 *** NSPlugin Wrapper *** handle_NPN_InvalidateRect *** NSPlugin Wrapper *** NPN_InvalidateRect instance=0xa847750 *** NSPlugin Wrapper *** done *** NSPlugin Viewer *** done *** NSPlugin Viewer *** handle_NPP_DestroyStream *** NSPlugin Viewer *** NPP_DestroyStream instance=0x8892830, stream=0x88ca118, reason=NPRES_DONE *** NSPlugin Viewer *** return: 0 [NPERR_NO_ERROR] *** NSPlugin Wrapper *** return: 0 [NPERR_NO_ERROR] *** NSPlugin Wrapper *** NPP_URLNotify instance=0xa847750, url='http://i4.ytimg.com/vi/SXyxy2l-vLQ/default.jpg', reason=0, notifyData=0x1f *** NSPlugin Wrapper *** done *** NSPlugin Wrapper *** NPP_SetWindow instance=0xa847750 *** NSPlugin Viewer *** handle_NPP_SetWindow *** NSPlugin Viewer *** NPP_SetWindow instance=0x8892830, window=(nil) *** NSPlugin Viewer *** return: 0 [NPERR_NO_ERROR] *** NSPlugin Wrapper *** return: 0 [NPERR_NO_ERROR] *** NSPlugin Wrapper *** NPP_HandleEvent instance=0xa847750 *** NSPlugin Viewer *** NPP_HandleEvent instance=0x8892830, event=0xbfa191a8 *** NSPlugin Viewer *** return: 0 *** NSPlugin Wrapper *** return: ret *** NSPlugin Wrapper *** NPP_Destroy instance=0xa847750 *** NSPlugin Viewer *** NPP_Destroy instance=0x8892830 *** NSPlugin Viewer *** NPN_ReleaseObject 0xb56f1200 *** NSPlugin Viewer *** return: 1 *** NSPlugin Viewer *** return: 0 [NPERR_NO_ERROR] *** NSPlugin Wrapper *** return: 0 [NPERR_NO_ERROR] *** NSPlugin Viewer *** NPClass::Invalidate(npobj 0xb56f1200) *** NSPlugin Viewer *** done *** NSPlugin Wrapper *** NPP_Destroy instance=0xab24ba0 *** NSPlugin Viewer *** NPP_Destroy instance=0x88c55b8 *** NSPlugin Viewer *** NPN_ReleaseObject 0xb56f1318 *** NSPlugin Viewer *** return: 1 *** NSPlugin Viewer *** return: 0 [NPERR_NO_ERROR] *** NSPlugin Wrapper *** return: 0 [NPERR_NO_ERROR] *** NSPlugin Viewer *** NPClass::Invalidate(npobj 0xb56f1318) *** NSPlugin Viewer *** done *** NSPlugin Wrapper *** NP_Shutdown *** NSPlugin Viewer *** handle_NP_Shutdown *** NSPlugin Viewer *** NP_Shutdown *** NSPlugin Viewer *** done *** NSPlugin Wrapper *** plugin_exit *** NSPlugin Viewer *** --- EXIT --- *** NSPlugin Wrapper *** return: 0 [NPERR_NO_ERROR] *** NSPlugin Wrapper *** NP_GetMIMEDescription *** NSPlugin Wrapper *** plugin_init for /usr/lib/flashplugin-nonfree/libflashplayer.so *** NSPlugin Wrapper *** NP_GetMIMEDescription *** NSPlugin Wrapper *** NP_GetMIMEDescription *** NSPlugin Wrapper *** NP_Initialize *** NSPlugin Wrapper *** NP_GetMIMEDescription *** NSPlugin Wrapper *** NP_GetMIMEDescription *** NSPlugin Wrapper *** NP_GetMIMEDescription *** NSPlugin Wrapper *** NP_Initialize *** NSPlugin Wrapper *** NP_Initialize *** NSPlugin Wrapper *** NP_GetMIMEDescription *** NSPlugin Wrapper *** NP_GetMIMEDescription *** NSPlugin Wrapper *** NP_GetMIMEDescription *** NSPlugin Wrapper *** NP_Initialize - Alexander From stransky at redhat.com Mon Oct 20 10:22:42 2008 From: stransky at redhat.com (Martin Stransky) Date: Mon, 20 Oct 2008 12:22:42 +0200 Subject: nspluginwrapper-1.1.2 invisible video In-Reply-To: <48FBBFAB.6070003@redhat.com> References: <48FBBFAB.6070003@redhat.com> Message-ID: <48FC5BF2.2060207@redhat.com> In fedora it could be caused by our sleep patch, it holds some X events and causes problem with adobe pdf plugin, too. Warren Togami wrote: > flash-plugin-10.0.12.36-release.i386 > nspluginwrapper-1.1.2-1.fc9.i386 > nspluginwrapper-1.1.2-1.fc9.x86_64 > firefox-3.0.2-1.fc9.x86_64 > > http://www.theonion.com/content/video/john_mccain_accidentally_left_on > If you reload a few times (or sometimes the first time), it somehow gets > into a state where you can hear the audio playing, but there is a blank > white area where the Flash video should be. Right-clicking the area > where the video should be shows an outline of the video, but no menus or > anything. Reloading the page fails to recover the video view. > > [root at newcaprica ~]# cat /etc/adobe/mms.cfg > WindowlessDisable=true > > Disabling WMODE in the plugin seems to allow the video to play. > > Warren Togami > wtogami at redhat.com > > _______________________________________________ > Nspluginwrapper-devel-list mailing list > Nspluginwrapper-devel-list at redhat.com > https://www.redhat.com/mailman/listinfo/nspluginwrapper-devel-list From webmaster at kwokinator.com Tue Oct 21 00:37:20 2008 From: webmaster at kwokinator.com (Adrien Kwok) Date: Mon, 20 Oct 2008 20:37:20 -0400 Subject: Can't compile 1.1.2 Message-ID: <1592de200810201737w7a58d09k872863fd22119cdb@mail.gmail.com> I got this error when trying to compile the software: gcc -std=c99 -m32 -o npviewer.bin npviewer-npw-viewer.o npviewer-npw-rpc.o npviewer-rpc.o npviewer-debug.o npviewer-utils.o npviewer-npruntime.o npviewer-cxxabi-compat.o -m32 -Llsb-build-i386 -lgtk-x11-2.0 -lgdk-x11-2.0 -lgobject-2.0 -ldl -lglib-2.0 -lX11 -lXt -ldl -lpthread -lgthread-2.0 -Wl,--export-dynamic -Wl,--version-script,/home/adrien/unzipped/nspluginwrapper-1.1.2/src/npw-viewer.map -lsupc++ /usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/bin/ld: skipping incompatible /usr/lib64/gcc/x86_64-suse-linux/4.3/libsupc++.a when searching for -lsupc++ /usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/bin/ld: cannot find -lsupc++ collect2: ld returned 1 exit status make: *** [npviewer.bin] Error 1 Any help is appreciated. Adrien -- Registered Linux User: 362616 Webmaster of Kwokinator.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From stransky at redhat.com Tue Oct 21 06:09:53 2008 From: stransky at redhat.com (Martin Stransky) Date: Tue, 21 Oct 2008 08:09:53 +0200 Subject: Can't compile 1.1.2 In-Reply-To: <1592de200810201737w7a58d09k872863fd22119cdb@mail.gmail.com> References: <1592de200810201737w7a58d09k872863fd22119cdb@mail.gmail.com> Message-ID: <48FD7231.1010107@redhat.com> You can't link 640bit library (libsupc++.a) with 32-bit binaries...your cross-compiler environment seems to be broken. Adrien Kwok wrote: > I got this error when trying to compile the software: > > gcc -std=c99 -m32 -o npviewer.bin npviewer-npw-viewer.o npviewer-npw-rpc.o > npviewer-rpc.o npviewer-debug.o npviewer-utils.o npviewer-npruntime.o > npviewer-cxxabi-compat.o -m32 -Llsb-build-i386 -lgtk-x11-2.0 -lgdk-x11-2.0 > -lgobject-2.0 -ldl -lglib-2.0 -lX11 -lXt -ldl -lpthread -lgthread-2.0 > -Wl,--export-dynamic > -Wl,--version-script,/home/adrien/unzipped/nspluginwrapper-1.1.2/src/npw-viewer.map > -lsupc++ > /usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/bin/ld: > skipping incompatible /usr/lib64/gcc/x86_64-suse-linux/4.3/libsupc++.a when > searching for -lsupc++ > /usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/bin/ld: > cannot find -lsupc++ > collect2: ld returned 1 exit status > make: *** [npviewer.bin] Error 1 > > Any help is appreciated. > > Adrien > > > ------------------------------------------------------------------------ > > _______________________________________________ > Nspluginwrapper-devel-list mailing list > Nspluginwrapper-devel-list at redhat.com > https://www.redhat.com/mailman/listinfo/nspluginwrapper-devel-list From webmaster at kwokinator.com Tue Oct 21 06:22:50 2008 From: webmaster at kwokinator.com (Adrien Kwok) Date: Tue, 21 Oct 2008 02:22:50 -0400 Subject: Can't compile 1.1.2 In-Reply-To: <48FD7231.1010107@redhat.com> References: <1592de200810201737w7a58d09k872863fd22119cdb@mail.gmail.com> <48FD7231.1010107@redhat.com> Message-ID: <1592de200810202322t790a29d6g2f546f004728571f@mail.gmail.com> So how do I go and fix this problem? I don't exactly know how to go about it. This is my configuration when I hit ./configure Install prefix /usr nspluginwrapper root dir /usr/lib/nspluginwrapper Strip binaries no Bi-arch build yes Build viewer yes Build for Linux only yes Build standalone player yes 32-bit library dir name lib 64-bit library dir name lib64 Source path /home/adrien/unzipped/nspluginwrapper-1.1.2 C compiler gcc -std=c99 C++ compiler g++ host OS linux host CPU x86_64 host big endian no target OS linux target CPU i386 rpc init timeout 5 secs Shouldn't the target CPU be x86_64 and not i386? Thanks, Adrien On Tue, Oct 21, 2008 at 2:09 AM, Martin Stransky wrote: > You can't link 640bit library (libsupc++.a) with 32-bit binaries...your > cross-compiler environment seems to be broken. > > Adrien Kwok wrote: > >> I got this error when trying to compile the software: >> >> gcc -std=c99 -m32 -o npviewer.bin npviewer-npw-viewer.o npviewer-npw-rpc.o >> npviewer-rpc.o npviewer-debug.o npviewer-utils.o npviewer-npruntime.o >> npviewer-cxxabi-compat.o -m32 -Llsb-build-i386 -lgtk-x11-2.0 -lgdk-x11-2.0 >> -lgobject-2.0 -ldl -lglib-2.0 -lX11 -lXt -ldl -lpthread -lgthread-2.0 >> -Wl,--export-dynamic >> >> -Wl,--version-script,/home/adrien/unzipped/nspluginwrapper-1.1.2/src/npw-viewer.map >> -lsupc++ >> /usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/bin/ld: >> skipping incompatible /usr/lib64/gcc/x86_64-suse-linux/4.3/libsupc++.a >> when >> searching for -lsupc++ >> /usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/bin/ld: >> cannot find -lsupc++ >> collect2: ld returned 1 exit status >> make: *** [npviewer.bin] Error 1 >> >> Any help is appreciated. >> >> Adrien >> >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Nspluginwrapper-devel-list mailing list >> Nspluginwrapper-devel-list at redhat.com >> https://www.redhat.com/mailman/listinfo/nspluginwrapper-devel-list >> > > _______________________________________________ > Nspluginwrapper-devel-list mailing list > Nspluginwrapper-devel-list at redhat.com > https://www.redhat.com/mailman/listinfo/nspluginwrapper-devel-list > -- Registered Linux User: 362616 Webmaster of Kwokinator.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From demontager at gmail.com Tue Oct 21 12:08:51 2008 From: demontager at gmail.com (Andrey) Date: Tue, 21 Oct 2008 10:08:51 -0200 Subject: Can't compile 1.1.2 In-Reply-To: <1592de200810202322t790a29d6g2f546f004728571f@mail.gmail.com> References: <1592de200810201737w7a58d09k872863fd22119cdb@mail.gmail.com> <48FD7231.1010107@redhat.com> <1592de200810202322t790a29d6g2f546f004728571f@mail.gmail.com> Message-ID: <48FDC653.6070604@gmail.com> Adrien Kwok ?????: > So how do I go and fix this problem? I don't exactly know how to go > about it. > > This is my configuration when I hit ./configure > > Install prefix /usr > nspluginwrapper root dir /usr/lib/nspluginwrapper > Strip binaries no > Bi-arch build yes > Build viewer yes > Build for Linux only yes > Build standalone player yes > 32-bit library dir name lib > 64-bit library dir name lib64 > Source path /home/adrien/unzipped/nspluginwrapper-1.1.2 > C compiler gcc -std=c99 > C++ compiler g++ > host OS linux > host CPU x86_64 > host big endian no > target OS linux > target CPU i386 > rpc init timeout 5 secs > > Shouldn't the target CPU be x86_64 and not i386? > > Thanks, > Adrien > > On Tue, Oct 21, 2008 at 2:09 AM, Martin Stransky > wrote: > > You can't link 640bit library (libsupc++.a) with 32-bit > binaries...your cross-compiler environment seems to be broken. > > Adrien Kwok wrote: > > I got this error when trying to compile the software: > > gcc -std=c99 -m32 -o npviewer.bin npviewer-npw-viewer.o > npviewer-npw-rpc.o > npviewer-rpc.o npviewer-debug.o npviewer-utils.o > npviewer-npruntime.o > npviewer-cxxabi-compat.o -m32 -Llsb-build-i386 -lgtk-x11-2.0 > -lgdk-x11-2.0 > -lgobject-2.0 -ldl -lglib-2.0 -lX11 -lXt -ldl -lpthread > -lgthread-2.0 > -Wl,--export-dynamic > -Wl,--version-script,/home/adrien/unzipped/nspluginwrapper-1.1.2/src/npw-viewer.map > -lsupc++ > /usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/bin/ld: > skipping incompatible > /usr/lib64/gcc/x86_64-suse-linux/4.3/libsupc++.a when > searching for -lsupc++ > /usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/bin/ld: > cannot find -lsupc++ > collect2: ld returned 1 exit status > make: *** [npviewer.bin] Error 1 > > Any help is appreciated. > > Adrien > > > ------------------------------------------------------------------------ > > _______________________________________________ > Nspluginwrapper-devel-list mailing list > Nspluginwrapper-devel-list at redhat.com > > https://www.redhat.com/mailman/listinfo/nspluginwrapper-devel-list > > > _______________________________________________ > Nspluginwrapper-devel-list mailing list > Nspluginwrapper-devel-list at redhat.com > > https://www.redhat.com/mailman/listinfo/nspluginwrapper-devel-list > > > > > -- > Registered Linux User: 362616 > Webmaster of Kwokinator.com > ------------------------------------------------------------------------ > Have a same problem like you, You need i saw somewhere that libstdc ++ i386 needed for this, i mean -lsupc++ > _______________________________________________ > Nspluginwrapper-devel-list mailing list > Nspluginwrapper-devel-list at redhat.com > https://www.redhat.com/mailman/listinfo/nspluginwrapper-devel-list From stransky at redhat.com Tue Oct 21 13:05:38 2008 From: stransky at redhat.com (Martin Stransky) Date: Tue, 21 Oct 2008 15:05:38 +0200 Subject: event patch Message-ID: <48FDD3A2.8070802@redhat.com> Hi Gwenole, I'm going to remove the sleep patch from fedora for now. But I'm going to use a different patch - it's attached. It removes the limit of processed events from xt_event_polling_timer_callback handler. I've tested it with flash 9&10 and pdf plugin. CPU usage was the same as w/o this patch but plug-ins (from my impression) behave more smoothly. Especially if I switched among different tabs/apps. Another thing comes to my mind - the runtime-restart patch has one side-effect what can cause browser crash. If the plugin crashes browser sometimes still tries to sens commands to it (because browser doesn't know the current plugin instance is actually dead). It doesn't matter for calls via. RPC but it's a problem when browser and plugin are connected via. npruntime (npclass_bridge), because pointers to already released memory (the handlers) are stored in browser internal NPObject and it still calls them... If you're interested I'll derive a minimal patch from the fedora one. Our restart patch checks the npruntime objects, rpc connections, adds error callbacks to the RPC code which hopefully terminates plug-ins w/o connected browser (what can happen when browser segfaults but wrapped plug-ins are still alive). ma. -------------- next part -------------- A non-text attachment was scrubbed... Name: nspluginwrapper-1.1.12-event.patch Type: text/x-patch Size: 583 bytes Desc: not available URL: From gb.public at free.fr Tue Oct 21 22:31:22 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Wed, 22 Oct 2008 00:31:22 +0200 Subject: Another testcase that still fails for me with the latest... In-Reply-To: References: <48FA1BD7.7090107@gmail.com> Message-ID: <601FA94C-CCA4-4F4B-B1F8-DF3B0CE327E4@free.fr> Hi, Le 19 oct. 08 ? 23:32, Ketan Patel a ?crit : > I am still seeing crashes when trying to use the grandcenral: > http://www.grandcentral.com Could you please tell: - Your distribution - What browser and version - Flash Player plugin version, windowless enabled or not - Did this ever worked before? - Does it work correctly with a 32-bit browser? Anyway, I could reproduce the problem on Fedora 9 x86_64 with the following packages: flash-plugin-10.0.12.36-release.i386 firefox-3.0.2-1.fc9.x86_64 Though it works correctly on Mandriva Linux LE2005 x86_64 with flash 10.0.0.525 firefox 1.0.2 | 2.0 The flash content appears to always be windowed anyway (no wmode= option as far as I can see). The viewer doesn't crash. Actually, it complains about gtk_style_detach() errors and such, which doesn't happen on Mandriva. That shouldn't be a problem though. Needs more testing as there are too many differences with those systems. Regards, Gwenole. From gb.public at free.fr Tue Oct 21 22:39:47 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Wed, 22 Oct 2008 00:39:47 +0200 Subject: Can't compile 1.1.2 In-Reply-To: <1592de200810201737w7a58d09k872863fd22119cdb@mail.gmail.com> References: <1592de200810201737w7a58d09k872863fd22119cdb@mail.gmail.com> Message-ID: Hi, Le 21 oct. 08 ? 02:37, Adrien Kwok a ?crit : > gcc -std=c99 -m32 -o npviewer.bin npviewer-npw-viewer.o npviewer-npw- > rpc.o npviewer-rpc.o npviewer-debug.o npviewer-utils.o npviewer- > npruntime.o npviewer-cxxabi-compat.o -m32 -Llsb-build-i386 -lgtk- > x11-2.0 -lgdk-x11-2.0 -lgobject-2.0 -ldl -lglib-2.0 -lX11 -lXt -ldl - > lpthread -lgthread-2.0 -Wl,--export-dynamic -Wl,--version-script,/ > home/adrien/unzipped/nspluginwrapper-1.1.2/src/npw-viewer.map -lsupc++ > /usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/ > bin/ld: skipping incompatible /usr/lib64/gcc/x86_64-suse-linux/4.3/ > libsupc++.a when searching for -lsupc++ > /usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/ > bin/ld: cannot find -lsupc++ > collect2: ld returned 1 exit status > make: *** [npviewer.bin] Error 1 You are missing 32bit packages. In particular, the following should help: libstdc++43-devel-32bit and others should be grabbed on the way. Regards, Gwenol?. From gb.public at free.fr Tue Oct 21 23:03:41 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Wed, 22 Oct 2008 01:03:41 +0200 Subject: nspluginwrapper-1.1.2 invisible video In-Reply-To: <48FBBFAB.6070003@redhat.com> References: <48FBBFAB.6070003@redhat.com> Message-ID: Hi, Le 20 oct. 08 ? 01:15, Warren Togami a ?crit : > http://www.theonion.com/content/video/john_mccain_accidentally_left_on > If you reload a few times (or sometimes the first time), it somehow > gets into a state where you can hear the audio playing, but there is > a blank white area where the Flash video should be. Right-clicking > the area where the video should be shows an outline of the video, > but no menus or anything. Reloading the page fails to recover the > video view. Thanks, I could reproduce the problem on Fedora 9 and an unpatched version of nspluginwrapper so the sleep patch is not in cause for this one. I have not investigated further yet but as far as I can see: the plugin is indeed running and in the event the video shows up, the NPP_SetWindow() happens earlier. Regards, Gwenole. From webmaster at kwokinator.com Wed Oct 22 05:22:53 2008 From: webmaster at kwokinator.com (Adrien Kwok) Date: Wed, 22 Oct 2008 01:22:53 -0400 Subject: Can't compile 1.1.2 In-Reply-To: References: <1592de200810201737w7a58d09k872863fd22119cdb@mail.gmail.com> Message-ID: <1592de200810212222m7c8c7de1w172ff15749471138@mail.gmail.com> Thanks that did it. Adrien On Tue, Oct 21, 2008 at 6:39 PM, Gwenole Beauchesne wrote: > Hi, > > Le 21 oct. 08 ? 02:37, Adrien Kwok a ?crit : > > gcc -std=c99 -m32 -o npviewer.bin npviewer-npw-viewer.o npviewer-npw-rpc.o >> npviewer-rpc.o npviewer-debug.o npviewer-utils.o npviewer-npruntime.o >> npviewer-cxxabi-compat.o -m32 -Llsb-build-i386 -lgtk-x11-2.0 -lgdk-x11-2.0 >> -lgobject-2.0 -ldl -lglib-2.0 -lX11 -lXt -ldl -lpthread -lgthread-2.0 >> -Wl,--export-dynamic >> -Wl,--version-script,/home/adrien/unzipped/nspluginwrapper-1.1.2/src/npw-viewer.map >> -lsupc++ >> /usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/bin/ld: >> skipping incompatible /usr/lib64/gcc/x86_64-suse-linux/4.3/libsupc++.a when >> searching for -lsupc++ >> /usr/lib64/gcc/x86_64-suse-linux/4.3/../../../../x86_64-suse-linux/bin/ld: >> cannot find -lsupc++ >> collect2: ld returned 1 exit status >> make: *** [npviewer.bin] Error 1 >> > > You are missing 32bit packages. In particular, the following should help: > libstdc++43-devel-32bit and others should be grabbed on the way. > > Regards, > Gwenol?. > > > _______________________________________________ > Nspluginwrapper-devel-list mailing list > Nspluginwrapper-devel-list at redhat.com > https://www.redhat.com/mailman/listinfo/nspluginwrapper-devel-list > -- Registered Linux User: 362616 Webmaster of Kwokinator.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From stransky at redhat.com Wed Oct 22 07:14:32 2008 From: stransky at redhat.com (Martin Stransky) Date: Wed, 22 Oct 2008 09:14:32 +0200 Subject: nspluginwrapper-1.1.2 invisible video In-Reply-To: References: <48FBBFAB.6070003@redhat.com> Message-ID: <48FED2D8.1000806@redhat.com> I can't reproduce it with latest fedora package with the event patch I posted yesterday... Gwenole Beauchesne wrote: > Hi, > > Le 20 oct. 08 ? 01:15, Warren Togami a ?crit : > >> http://www.theonion.com/content/video/john_mccain_accidentally_left_on >> If you reload a few times (or sometimes the first time), it somehow >> gets into a state where you can hear the audio playing, but there is a >> blank white area where the Flash video should be. Right-clicking the >> area where the video should be shows an outline of the video, but no >> menus or anything. Reloading the page fails to recover the video view. > > Thanks, I could reproduce the problem on Fedora 9 and an unpatched > version of nspluginwrapper so the sleep patch is not in cause for this > one. I have not investigated further yet but as far as I can see: the > plugin is indeed running and in the event the video shows up, the > NPP_SetWindow() happens earlier. > > Regards, > Gwenole. > > _______________________________________________ > Nspluginwrapper-devel-list mailing list > Nspluginwrapper-devel-list at redhat.com > https://www.redhat.com/mailman/listinfo/nspluginwrapper-devel-list From ktp420 at live.com Wed Oct 22 13:15:39 2008 From: ktp420 at live.com (Ketan Patel) Date: Wed, 22 Oct 2008 09:15:39 -0400 Subject: Another testcase that still fails for me with the latest... In-Reply-To: <601FA94C-CCA4-4F4B-B1F8-DF3B0CE327E4@free.fr> References: <48FA1BD7.7090107@gmail.com> <601FA94C-CCA4-4F4B-B1F8-DF3B0CE327E4@free.fr> Message-ID: I am on Ubuntu 8.10 amd64 with following: firefox 3.0.3+nobinonly-0ubuntu2 flashplugin-nonfree 10.0.12.36ubuntu1 I also noticed following errors: (npviewer.bin:6576): Gtk-CRITICAL **: gtk_widget_destroy: assertion `GTK_IS_WIDGET (widget)' failed This was working pretty good before upgrading the nspluginwrapper. ---------------------------------------- > From: gb.public at free.fr > To: nspluginwrapper-devel-list at redhat.com > Date: Wed, 22 Oct 2008 00:31:22 +0200 > Subject: Re: Another testcase that still fails for me with the latest... > > Hi, > > Le 19 oct. 08 ? 23:32, Ketan Patel a ?crit : > >> I am still seeing crashes when trying to use the grandcenral: >> http://www.grandcentral.com > > Could you please tell: > - Your distribution > - What browser and version > - Flash Player plugin version, windowless enabled or not > - Did this ever worked before? > - Does it work correctly with a 32-bit browser? > > Anyway, I could reproduce the problem on Fedora 9 x86_64 with the > following packages: > flash-plugin-10.0.12.36-release.i386 > firefox-3.0.2-1.fc9.x86_64 > > Though it works correctly on Mandriva Linux LE2005 x86_64 with > flash 10.0.0.525 > firefox 1.0.2 | 2.0 > > The flash content appears to always be windowed anyway (no wmode= > option as far as I can see). > > The viewer doesn't crash. Actually, it complains about > gtk_style_detach() errors and such, which doesn't happen on Mandriva. > That shouldn't be a problem though. Needs more testing as there are > too many differences with those systems. > > Regards, > Gwenole. > > _______________________________________________ > Nspluginwrapper-devel-list mailing list > Nspluginwrapper-devel-list at redhat.com > https://www.redhat.com/mailman/listinfo/nspluginwrapper-devel-list _________________________________________________________________ Want to read Hotmail messages in Outlook? The Wordsmiths show you how. http://windowslive.com/connect/post/wedowindowslive.spaces.live.com-Blog-cns!20EE04FBC541789!167.entry?ocid=TXT_TAGLM_WL_hotmail_092008 From gb.public at free.fr Fri Oct 24 05:04:04 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Fri, 24 Oct 2008 07:04:04 +0200 Subject: nspluginwrapper-1.1.2 invisible video In-Reply-To: <48FED2D8.1000806@redhat.com> References: <48FBBFAB.6070003@redhat.com> <48FED2D8.1000806@redhat.com> Message-ID: <2D4D59F0-C4DE-4F51-8EE4-6ADD48FA10D9@free.fr> Hi, Le 22 oct. 08 ? 09:14, Martin Stransky a ?crit : > I can't reproduce it with latest fedora package with the event > patch I posted yesterday... I tried the trick (among others) and it didn't work for me. So, I'd suspect a synchronisation issue in the init phase (from NPP_SetWindow()) or some uninitialized (not zero'ed?) variable. This is strange because the plugin receives the stream normally and still emits NPN_InvalidateRect() with reasonable values. I have not explored the offscreen pixmap yet but I think it is correct too. The plugin probably received the browser GraphicsExpose event too early and disregarded it. In general, sources of the plugin would help to know in this case. ;-) Ketan's problem looks the same (I mean, visually) but I am inclined to think they are different. Here, it would be a windowless specific problem whereas Ketan's would be an NPObject one. I definitely prefer when it crashes. Regards, Gwenole. From gb.public at free.fr Tue Oct 28 08:10:15 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Tue, 28 Oct 2008 09:10:15 +0100 (CET) Subject: Another testcase that still fails for me with the latest... In-Reply-To: <27698708.429921225181310489.JavaMail.root@spooler5-g27.priv.proxad.net> Message-ID: <5669592.430201225181415338.JavaMail.root@spooler5-g27.priv.proxad.net> Hi, ----- "Ketan Patel" a ?crit : > I am on Ubuntu 8.10 amd64 with following: > firefox 3.0.3+nobinonly-0ubuntu2 > flashplugin-nonfree 10.0.12.36ubuntu1 OK, the common things are Flash 10 and Firefox 3. Actually, I tried the following browsers and they don't exhibit this behaviour: Firefox up to 2.0.0.17 and WebKit from last week. In the WebKit case, the plugin window can be empty for up to 1 or 2 seconds though, but it always shows up. Then, I tried to backport some of the NPAPI changes from Firefox 3 to Firefox 2, mostly related to windowed stuff. In particular, patches from BZ #386493, #386493, #384845 (modules/plugin part only), #26152 (likewise), #386844. No change, patched firefox 2 is still safe to the behaviour you describe. However, I think this could be related to fix for #1156. Not that the fix is in cause but it could exhaust the nspluginwrapper bug. It seems that, in failing cases, the plugin is not receiving the expected NPP_SetWindow(). Not 100% sure yet but I compared both failing and successful cases (for the same amount of page loads through the tabs, and same order) and that's a difference. > I also noticed following errors: > (npviewer.bin:6576): Gtk-CRITICAL **: gtk_widget_destroy: assertion > `GTK_IS_WIDGET (widget)' failed I have attached a fix I committed for this. The extra GtkPlug/GtkSocket allocation is a pure hack and it should work without, but Flash doesn't. Funny enough, Opera people did not bother to find a better solution either and apparently used the same hack as I... Though, I will probably try a correct fix later. > This was working pretty good before upgrading the nspluginwrapper. I could reproduce the problem with nspluginwrapper 1.0 too but I will try again now you raised the doubt. ;-) Regards, Gwenol?. -------------- next part -------------- A non-text attachment was scrubbed... Name: nspluginwrapper-avoid-double-free.patch Type: text/x-diff Size: 858 bytes Desc: not available URL: From gb.public at free.fr Wed Oct 29 05:53:21 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Wed, 29 Oct 2008 06:53:21 +0100 Subject: [PATCH] Fix NPP_URLNotify() Message-ID: <1AA4B73A-C0E2-471A-991D-5CED82C360D1@free.fr> Hi, Bennet Yee noticed this one liner. Committed to trunk and 1.0-branch. -------------- next part -------------- A non-text attachment was scrubbed... Name: nspluginwrapper-fix-urlnotify.patch Type: application/octet-stream Size: 502 bytes Desc: not available URL: -------------- next part -------------- From gb.public at free.fr Wed Oct 29 05:56:42 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Wed, 29 Oct 2008 06:56:42 +0100 Subject: [PATCH] Fix reported NPAPI interface Message-ID: <7552C1E5-10DD-4153-98CE-A1E466EC888C@free.fr> Hi, Since we currenly don't implement NPAPI >= 0.18, force it to 0.17 ATM. -------------- next part -------------- A non-text attachment was scrubbed... Name: nspluginwrapper-npapi-version.patch Type: application/octet-stream Size: 1636 bytes Desc: not available URL: -------------- next part -------------- From gb.public at free.fr Wed Oct 29 05:58:44 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Wed, 29 Oct 2008 06:58:44 +0100 Subject: Another testcase that still fails for me with the latest... In-Reply-To: <5669592.430201225181415338.JavaMail.root@spooler5-g27.priv.proxad.net> References: <5669592.430201225181415338.JavaMail.root@spooler5-g27.priv.proxad.net> Message-ID: Le 28 oct. 08 ? 09:10, Gwenole Beauchesne a ?crit : >> This was working pretty good before upgrading the nspluginwrapper. > > I could reproduce the problem with nspluginwrapper 1.0 too but I > will try again now you raised the doubt. ;-) Yes, I tried again this morning, and it does fail with nspluginwrapper 1.0 too. Ketan, what version worked for you? Thanks. From gb.public at free.fr Wed Oct 29 06:07:14 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Wed, 29 Oct 2008 07:07:14 +0100 Subject: [PATCH] Fix reported NPAPI interface In-Reply-To: <7552C1E5-10DD-4153-98CE-A1E466EC888C@free.fr> References: <7552C1E5-10DD-4153-98CE-A1E466EC888C@free.fr> Message-ID: Le 29 oct. 08 ? 06:56, Gwenole Beauchesne a ?crit : > Since we currenly don't implement NPAPI >= 0.18, force it to 0.17 ATM. I forgot to mention this one applies to current trunk but you may just want to define version to 17 and set NPClass::version directly to 1. This is only important for plugins that would use NPN_Enumerate() et al. To my knowledge this is only useful to NPAPI Java plugins for now and they don't work yet anyway. i.e. don't bother urging the fix for your soon-to-be-released distributions. ;-) From gb.public at free.fr Wed Oct 29 06:35:42 2008 From: gb.public at free.fr (Gwenole Beauchesne) Date: Wed, 29 Oct 2008 07:35:42 +0100 Subject: event patch In-Reply-To: <48FDD3A2.8070802@redhat.com> References: <48FDD3A2.8070802@redhat.com> Message-ID: <6315FE3E-7D13-4B08-8371-4B6CBA8926C1@free.fr> Hi Martin, Le 21 oct. 08 ? 15:05, Martin Stransky a ?crit : > I'm going to remove the sleep patch from fedora for now. But I'm > going to use a different patch - it's attached. It removes the limit > of processed events from xt_event_polling_timer_callback handler. For the records, I have attached a patch I already sent you I think but there is still one problem with it. Basically, the old sleep patch introduced two regressions in older plugins: one with jpeg2k plugin, another one with djvulibre. The former was due to an Xt timeout. The other is relevant for an Xt input. Unfortunately, run-time detecting internal Xt structs for that is a nightmare. So, removing the patch is a current better alternative. ;-) > I've tested it with flash 9&10 and pdf plugin. CPU usage was the > same as w/o this patch but plug-ins (from my impression) behave > more smoothly. Especially if I switched among different tabs/apps. That's a good thing I think. If the timeout handler calls into the browser, it should work now that RPC is concurrent aware. However, it does not resolve our other issues yet but makes them less frequent. I will work on that case again later. > It doesn't matter for calls via. RPC but it's a problem when browser > and plugin are connected via. npruntime (npclass_bridge), because > pointers to already released memory (the handlers) are stored in > browser internal NPObject and it still calls them... Talking about using released memory in NPObject (or others), I wonder if someone already ran npviewer.bin with some MALLOC_PERTURB_ set. I think there are cases where we reuse memory that was just released. I will try this evening. It should be an interesting intermediate test before valgrind'ing (it reports too many problems not necessarily caused by npviewer). Regards, Gwenole. -------------- next part -------------- A non-text attachment was scrubbed... Name: npw.sleep.diff Type: application/octet-stream Size: 4841 bytes Desc: not available URL: -------------- next part -------------- From ktp420 at live.com Fri Oct 31 02:01:23 2008 From: ktp420 at live.com (Ketan Patel) Date: Thu, 30 Oct 2008 22:01:23 -0400 Subject: Another testcase that still fails for me with the latest... In-Reply-To: References: <5669592.430201225181415338.JavaMail.root@spooler5-g27.priv.proxad.net> Message-ID: I was using following ubuntu version: https://launchpad.net/ubuntu/+source/nspluginwrapper/1.1.0-0ubuntu1 I also had following in my /etc/adobe/mms.cfg WindowlessDisable=true #OverrideGPUValidation=true I was using the flash 10 rc1 at the time. And firefox might have been at 3.1. Right now the firefox is at 3.0.3. Lots of stuff changed on the system so could be anything. ---------------------------------------- > From: gb.public at free.fr > To: nspluginwrapper-devel-list at redhat.com > Date: Wed, 29 Oct 2008 06:58:44 +0100 > Subject: Re: Another testcase that still fails for me with the latest... > > Le 28 oct. 08 ? 09:10, Gwenole Beauchesne a ?crit : > >>> This was working pretty good before upgrading the nspluginwrapper. >> >> I could reproduce the problem with nspluginwrapper 1.0 too but I >> will try again now you raised the doubt. ;-) > > Yes, I tried again this morning, and it does fail with nspluginwrapper > 1.0 too. Ketan, what version worked for you? Thanks. > > _______________________________________________ > Nspluginwrapper-devel-list mailing list > Nspluginwrapper-devel-list at redhat.com > https://www.redhat.com/mailman/listinfo/nspluginwrapper-devel-list _________________________________________________________________ You live life beyond your PC. So now Windows goes beyond your PC. http://clk.atdmt.com/MRT/go/115298556/direct/01/ From mtrainer at central-data.net Fri Oct 31 05:26:21 2008 From: mtrainer at central-data.net (Murray Trainer) Date: Fri, 31 Oct 2008 13:26:21 +0800 (GMT+08:00) Subject: Citrix ICA plugin on Ubuntu Hardy In-Reply-To: <13098367.815121225430587436.JavaMail.root@mailstore01.gopc.net> Message-ID: <9915203.815211225430781775.JavaMail.root@mailstore01.gopc.net> Hi, I had the Citrix ICA plugin working fine on Ubuntu Gutsy (7.10) 64-bit using Firefox 2 and the default nspluginwrapper on Gutsy (0.9.x from memory). Now I have upgraded to Ubuntu Hardy (8.04) 64-bit and Firefox 3.0.3. The ICA plugin installs and lists fine with nspluginwrapper but doesn't appear in the about:plugins list of plugins in Firefox 3.0.3. I have tried nspluginwrapper 0.9.91 and 1.0.0-1 with the same problem. Is this a known issue with Firefox 3? Any suggestions on a fix or degugging the problem? Thanks Murray Trainer -------------- next part -------------- An HTML attachment was scrubbed... URL: