59183 2003-05-31 19:25:02 +0000 arts_suspend() returns 0 if arts doesn't have DSP device open 2004-02-22 02:26:53 +0000 1 1 10 Unmaintained arts general 1.1.1 Mandrake RPMs Linux RESOLVED FIXED NOR normal --- 1 luigiwalser staikos 0 oldest_to_newest 136957 0 luigiwalser 2003-05-31 19:25:02 +0000 Version: 1.1.1 (using KDE KDE 3.1.1) Installed from: Mandrake RPMs Compiler: gcc version 3.2.2 (Mandrake Linux 9.2 3.2.2-5mdk) OS: Linux arts_suspend() will have artsd let go of the DSP device if it has it open. If it's sucessful, the postcondition is that artsd doesn't have the DSP device open, and 1 is returned. There is one condition where this postcondition is broken, that's when you call arts_suspend() and artsd didn't have the DSP device open in the first place. As far as trying to get artsd to let go of the device, it shouldn't matter whether it had it open in the first place or not, if arts can manage to make sure it's not holding the device (whether it needs to do any work to make that happen or not), 1 should be returned. 0 should only be returned when, as it said in the header file, "there were active non-suspendable modules," meaning artsd has the device open and is not going to give it up. arts_suspend() should not return 0 when arts doesn't have the DSP device open. 142623 1 staikos 2003-06-27 09:32:39 +0000 This looks very easy to fix, but can you tell me what you think it should return in this case, and provide a testcase to verify the results? I don't want to commit a fix without this information. Thanks 142658 2 luigiwalser 2003-06-27 12:03:01 +0000 Subject: Re: arts_suspend() returns 0 if arts doesn't have DSP device open Hi, I apologize for e-mailing you directly, but KDE Bugzilla is broken at the moment. When I try to submit an update it says "product was not defined; this may indicate a bug in your browser." It says that whether I'm in Konqueror or Mozilla. I'll paste what I was trying to write here (and attach the mentioned patch to this e-mail). Thank you. In the header file, it says "asks aRtsd to free the DSP device and return 1 if it was successful," so as long as artsd isn't holding the device open after the function is run, 1 should be returned. Here's a testcase: killall artsd artsd -s 45 artsdsp play $(dirname $(dirname `which startkde`))/share/sounds/KDE_Startup.wav Then wait one minute, and run a piece of C code that runs arts_suspend(). It should return 1. If you want some code, if you're running RedHat or Mandrake, get the newest TiMidity++ source RPM package from RedHat Rawhide, or Mandrake Cooker, and replace the patch that's name ends in detect.patch with the one I attach here. If you want me to make you a tarball, let me know. When you compile and run timidity with this patch, it will use arts output if arts_suspend returns 0, and OSS output if it returns 1. The easiest way I can tell which output it's using is if it's using arts you see the infamous arts error message: mcop warning: user defined signal handler found for SIG_PIPE, overriding you could also just try playing something else with arts simultaneously. Anyway, the expected behavior with this patched timidity is this: arts is currently playing something, timidity uses arts output arts is not currently playing something (whether 45 seconds have passed and it's already given up the device or not), timidity uses OSS output --- George Staikos <staikos@kde.org> wrote: > ------- You are receiving this mail because: ------- > You reported the bug, or are watching the reporter. > > http://bugs.kde.org/show_bug.cgi?id=59183 > staikos@kde.org changed: > > What |Removed > |Added > ---------------------------------------------------------------------------- > Status|UNCONFIRMED |NEW > everconfirmed|0 |1 > > > > ------- Additional Comments From staikos@kde.org > 2003-06-27 09:32 ------- > This looks very easy to fix, but can you tell me > what you think it should > return in this case, and provide a testcase to > verify the results? I don't > want to commit a fix without this information. > > Thanks __________________________________ Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month! http://sbc.yahoo.com Created an attachment (id=1893) TiMidity++-2.12.0-pre1-detect.patch.luigi 143509 3 staikos 2003-07-02 18:00:07 +0000 Correct, it returns -2 right now. I'll repair asap. 143511 4 staikos 2003-07-02 18:08:20 +0000 sorry, correction: it returns 0, not -2 :) 143518 5 luigiwalser 2003-07-02 18:26:31 +0000 I was about to say :o) I think I found the problem. I'm looking now at arts/soundserver/soundserver_impl.cc in the function SoundServer_impl::suspend() and it looks like this currently: bool SoundServer_impl::suspend() { if(Dispatcher::the()->flowSystem()->suspendable() && !Dispatcher::the()->flowSystem()->suspended()) { Dispatcher::the()->flowSystem()->suspend(); arts_info("sound server suspended by client"); return true; } return false; } I believe it should look like the following: bool SoundServer_impl::suspend() { if(Dispatcher::the()->flowSystem()->suspended() return true; if(Dispatcher::the()->flowSystem()->suspendable() { Dispatcher::the()->flowSystem()->suspend(); arts_info("sound server suspended by client"); return true; } return false; } 143520 6 staikos 2003-07-02 18:29:28 +0000 Subject: Re: arts_suspend() returns 0 if arts doesn't have DSP device open On Wednesday 02 July 2003 12:26, you wrote: > ------- I was about to say :o) > > I think I found the problem. I'm looking now at > arts/soundserver/soundserver_impl.cc in the function > SoundServer_impl::suspend() and it looks like this currently: bool > SoundServer_impl::suspend() { > if(Dispatcher::the()->flowSystem()->suspendable() && > !Dispatcher::the()->flowSystem()->suspended()) > { > Dispatcher::the()->flowSystem()->suspend(); > arts_info("sound server suspended by client"); > return true; > } > return false; > } > > I believe it should look like the following: > bool SoundServer_impl::suspend() { > if(Dispatcher::the()->flowSystem()->suspended() > return true; > if(Dispatcher::the()->flowSystem()->suspendable() { > Dispatcher::the()->flowSystem()->suspend(); > arts_info("sound server suspended by client"); > return true; > } > return false; > } Hah I was *just* about to send the exact same thing back to you! Ok I will commit this fix as I believe it to be correct too. Thanks! 143522 7 staikos 2003-07-02 18:33:35 +0000 Subject: ARTS_1_1_BRANCH: arts/soundserver CVS commit by staikos: Fix 59183 with the excellent help of the reporter. This could use a forward port for those who have arts-head checked out. thanks. CCMAIL: 59183-done@bugs.kde.org M +5 -2 soundserver_impl.cc 1.7.4.2 --- arts/soundserver/soundserver_impl.cc #1.7.4.1:1.7.4.2 @@ -63,6 +63,9 @@ long SoundServer_impl::secondsUntilSuspe bool SoundServer_impl::suspend() { - if(Dispatcher::the()->flowSystem()->suspendable() && - !Dispatcher::the()->flowSystem()->suspended()) + if (Dispatcher::the()->flowSystem()->suspended()) + { + return true; + } + if(Dispatcher::the()->flowSystem()->suspendable()) { Dispatcher::the()->flowSystem()->suspend(); 208423 8 stefan.bruens 2004-02-22 02:26:53 +0000 This is a duplicate of BR 44566 1893 2003-06-27 12:03:01 +0000 2003-06-27 12:03:01 +0000 TiMidity++-2.12.0-pre1-detect.patch.luigi TiMidity++-2.12.0-pre1-detect.patch.luigi application/octet-stream 990 luigiwalser ZGlmZiAtdU5yIFRpTWlkaXR5KystMi4xMi4wLXByZTEub3JpZy90aW1pZGl0eS90aW1pZGl0eS5j IFRpTWlkaXR5KystMi4xMi4wLXByZTEvdGltaWRpdHkvdGltaWRpdHkuYwotLS0gVGlNaWRpdHkr Ky0yLjEyLjAtcHJlMS5vcmlnL3RpbWlkaXR5L3RpbWlkaXR5LmMJMjAwMi0wMS0xNSAyMjozNjoz MS4wMDAwMDAwMDAgLTA1MDAKKysrIFRpTWlkaXR5KystMi4xMi4wLXByZTEvdGltaWRpdHkvdGlt aWRpdHkuYwkyMDAzLTA1LTI5IDIwOjE5OjAwLjAwMDAwMDAwMCAtMDQwMApAQCAtMzMxNyw2ICsz MzE3LDcgQEAKICAgICBpbnQgbmZpbGVzOwogICAgIGNoYXIgKipmaWxlczsKICAgICBpbnQgbWFp bl9yZXQ7CisgICAgaW50IG1vZGVfc2V0PTA7CiAKICNpZiBkZWZpbmVkKERBTkdFUk9VU19SRU5J Q0UpICYmICFkZWZpbmVkKF9fVzMyX18pICYmICFkZWZpbmVkKG1haW4pCiAgICAgLyoKQEAgLTM0 MTMsNiArMzQxNCwyOSBAQAogICAgIH0KICNlbmRpZgogCisjaWZkZWYgQVVfQVJUUworICAgIGlm KGFydHNfaW5pdCgpPT0wKSB7CisJICAgIGlmKGFydHNfc3VzcGVuZCgpPT0wKSB7CisJCSAgICBz ZXRfcGxheV9tb2RlKCJrIik7CisJCSAgICBtb2RlX3NldD0xOworCSAgICB9CisJICAgIGFydHNf ZnJlZSgpOworICAgIH0KKyNlbmRpZgorI2lmZGVmIEFVX0VTRAorICAgIGlmKCFtb2RlX3NldCkg eworCSAgICBpZighYWNjZXNzKCIvdXNyL2xpYi9saWJlc2Quc28uMCIsIFJfT0spKSB7CisJCSAg ICBzZXRlbnYoIkVTRF9OT19TUEFXTiIsICIxIiwgMCk7CisJCSAgICBzZXRfcGxheV9tb2RlKCJl Iik7CisJCSAgICBtb2RlX3NldD0xOworCSAgICB9CisgICAgfQorI2VuZGlmCisjaWZkZWYgQVVf T1NTCisgICAgaWYoIW1vZGVfc2V0KQorCSAgICBzZXRfcGxheV9tb2RlKCJkIik7CisjZW5kaWYK KyAgICAKICAgICBpZigoZXJyID0gdGltaWRpdHlfcHJlX2xvYWRfY29uZmlndXJhdGlvbigpKSAh PSAwKQogCXJldHVybiBlcnI7CiAK