Bug#908449: firefox-esr_60.2.0esr-1~deb9u2.1_i386.deb not working
On 18/09/18 21:45, Mike Hommey wrote: Incidentally, since you're talking about Javascript, I thought it might be relevant that I'm using the NoScript add-on, so I thought I'd try disabling that (which worked), and setting javascript.options.baselinejit and javascript.options.ion back to true. When I switched javascript.options.baselinejit back to true, Firefox instantly crashed. (But, of course, it's recoverable again via -safe-mode.) Setting only javascript.options.ion back to true *doesn't* crash Firefox, or prevent it from starting. Can you restore all the prefs and addons you changed to what they were before and try the following package? https://people.debian.org/~glandium/firefox-esr_60.2.0esr-1~deb9u2.2_i386.deb sha256: 4f092e11faf729316ea96be5b9f6ecca58ea1f001279c44406a8d505b021c17f I re-enabled NoScript and set javascript.options.ion and javascript.options.baselinejit back to true; then I installed your latest package, and it seems to work fine! Thanks for all your patient work on this. Tim <><
Bug#908449: firefox-esr_60.2.0esr-1~deb9u2.1_i386.deb not working
On Tue, Sep 18, 2018 at 07:17:15PM +1200, Tim Makarios wrote: > On 18/09/18 17:55, Mike Hommey wrote: > > So the crash is happening in JITed code, which is not expected: the JIT > > should disable itself. There might be something wrong either in the > > detection or maybe some parts don't disable themselves as they're > > supposed to. > > > > Can you try starting with `firefox-esr -safe-mode`? > > That works. > > > If that works, can > > you then go to about:config and switch javascript.options.ion to false, > > then restart without -safe-mode? > > That crashes. > > > If that works, stop there. If it > > doesn't work, can you try -safe-mode again, and switch > > javascript.options.baselinejit to false? > > And now it works in non-safe-mode! > > Incidentally, since you're talking about Javascript, I thought it might be > relevant that I'm using the NoScript add-on, so I thought I'd try disabling > that (which worked), and setting javascript.options.baselinejit and > javascript.options.ion back to true. When I switched > javascript.options.baselinejit back to true, Firefox instantly crashed. > (But, of course, it's recoverable again via -safe-mode.) > > Setting only javascript.options.ion back to true *doesn't* crash Firefox, or > prevent it from starting. Can you restore all the prefs and addons you changed to what they were before and try the following package? https://people.debian.org/~glandium/firefox-esr_60.2.0esr-1~deb9u2.2_i386.deb sha256: 4f092e11faf729316ea96be5b9f6ecca58ea1f001279c44406a8d505b021c17f (Others who have experienced the crash on startup are invited to try this package as well) Thanks Mike
Bug#908449: firefox-esr_60.2.0esr-1~deb9u2.1_i386.deb not working
On 18/09/18 17:55, Mike Hommey wrote: So the crash is happening in JITed code, which is not expected: the JIT should disable itself. There might be something wrong either in the detection or maybe some parts don't disable themselves as they're supposed to. Can you try starting with `firefox-esr -safe-mode`? That works. If that works, can you then go to about:config and switch javascript.options.ion to false, then restart without -safe-mode? That crashes. If that works, stop there. If it doesn't work, can you try -safe-mode again, and switch javascript.options.baselinejit to false? And now it works in non-safe-mode! Incidentally, since you're talking about Javascript, I thought it might be relevant that I'm using the NoScript add-on, so I thought I'd try disabling that (which worked), and setting javascript.options.baselinejit and javascript.options.ion back to true. When I switched javascript.options.baselinejit back to true, Firefox instantly crashed. (But, of course, it's recoverable again via -safe-mode.) Setting only javascript.options.ion back to true *doesn't* crash Firefox, or prevent it from starting. Tim <><
Bug#908449: firefox-esr_60.2.0esr-1~deb9u2.1_i386.deb not working
On Tue, Sep 18, 2018 at 02:30:03PM +0900, Mike Hommey wrote: > On Tue, Sep 18, 2018 at 05:26:43PM +1200, Tim Makarios wrote: > > On 18/09/18 17:05, Mike Hommey wrote: > > > Oh, right. Forgot about the crash reporter interposing itself. You can > > > do one of the following: > > > ... > > > - or send the report to Mozilla, and find its id in > > >~/.mozilla/firefox/Crash Reports/submitted > > > ... > > > > I'd already reported a few earlier crashes to Mozilla, but I did it again > > today, and the most recent file in that directory contains just the line: > > > > Crash ID: bp-a47d7854-9ea2-4c6b-b432-4b98e0180918 > > > > Is that what you're looking for? > > Yes, thanks. So the crash is happening in JITed code, which is not expected: the JIT should disable itself. There might be something wrong either in the detection or maybe some parts don't disable themselves as they're supposed to. Can you try starting with `firefox-esr -safe-mode`? If that works, can you then go to about:config and switch javascript.options.ion to false, then restart without -safe-mode? If that works, stop there. If it doesn't work, can you try -safe-mode again, and switch javascript.options.baselinejit to false? Can you then report what worked and what didn't? Thanks Mike
Bug#908449: firefox-esr_60.2.0esr-1~deb9u2.1_i386.deb not working
On Tue, Sep 18, 2018 at 05:26:43PM +1200, Tim Makarios wrote: > On 18/09/18 17:05, Mike Hommey wrote: > > Oh, right. Forgot about the crash reporter interposing itself. You can > > do one of the following: > > ... > > - or send the report to Mozilla, and find its id in > >~/.mozilla/firefox/Crash Reports/submitted > > ... > > I'd already reported a few earlier crashes to Mozilla, but I did it again > today, and the most recent file in that directory contains just the line: > > Crash ID: bp-a47d7854-9ea2-4c6b-b432-4b98e0180918 > > Is that what you're looking for? Yes, thanks. Mike
Bug#908449: firefox-esr_60.2.0esr-1~deb9u2.1_i386.deb not working
On 18/09/18 17:05, Mike Hommey wrote: Oh, right. Forgot about the crash reporter interposing itself. You can do one of the following: ... - or send the report to Mozilla, and find its id in ~/.mozilla/firefox/Crash Reports/submitted ... I'd already reported a few earlier crashes to Mozilla, but I did it again today, and the most recent file in that directory contains just the line: Crash ID: bp-a47d7854-9ea2-4c6b-b432-4b98e0180918 Is that what you're looking for? Tim <><
Bug#908449: firefox-esr_60.2.0esr-1~deb9u2.1_i386.deb not working
On Tue, Sep 18, 2018 at 04:55:07PM +1200, Tim Makarios wrote: > On 13/09/18 21:23, Mike Hommey wrote: > > Huh. Can you then try ulimit -c unlimited, and upload the resulting core > > file somewhere? > > I'm not sure I quite understand you here. I ran > > $ ulimit -c unlimited > $ firefox-esr > > The output was: > > > ExceptionHandler::GenerateDump cloned child 1472 > > ExceptionHandler::SendContinueSignalToChild sent continue signal to child > > ExceptionHandler::WaitForContinueSignal waiting for continue signal... > > and a window popped up inviting me to choose whether to report the crash to > Mozilla, and whether to quit or restart Firefox. > > I couldn't find any resulting core file. Where is it meant to be? Oh, right. Forgot about the crash reporter interposing itself. You can do one of the following: - set MOZ_CRASHREPORTER_DISABLE to 1 in the environment and try again - or send the report to Mozilla, and find its id in ~/.mozilla/firefox/Crash Reports/submitted - or attach the crash report data from ~/.mozilla/firefox/Crash Reports/pending (both the .dmp and the .extra with the same id; only the last pair) I'd suggest the second one, in case your session crashed at some point where some private data might have been on the stack. That would limit the number of people who can look at the data, while attaching the raw data to the bug would make it public. Or the third one, if you send the files to me rather than the bug (that would also be smaller to transfer than a full core). Thanks Mike
Bug#908449: firefox-esr_60.2.0esr-1~deb9u2.1_i386.deb not working
On 13/09/18 21:23, Mike Hommey wrote: Huh. Can you then try ulimit -c unlimited, and upload the resulting core file somewhere? I'm not sure I quite understand you here. I ran $ ulimit -c unlimited $ firefox-esr The output was: ExceptionHandler::GenerateDump cloned child 1472 ExceptionHandler::SendContinueSignalToChild sent continue signal to child ExceptionHandler::WaitForContinueSignal waiting for continue signal... and a window popped up inviting me to choose whether to report the crash to Mozilla, and whether to quit or restart Firefox. I couldn't find any resulting core file. Where is it meant to be? Tim <><
Bug#908449: firefox-esr_60.2.0esr-1~deb9u2.1_i386.deb not working
On Thu, Sep 13, 2018 at 09:04:54PM +1200, Tim Makarios wrote: > On 13/09/18 18:42, Mike Hommey wrote: > > Can you install the firefox-esr-dbgsym package from > > https://people.debian.org/~glandium/firefox-esr-dbgsym_60.2.0esr-1~deb9u2.1_i386.deb > > > > and run the following command: > > > > gdb firefox-esr > > > > At the gdb prompt, type the following commands: > > > >set pagination off > >run > > When I get to this point, it gives the output: > > Starting program: /usr/bin/firefox-esr [Thread debugging using > > libthread_db enabled] > > Using host libthread_db library "/lib/i386-linux-gnu/libthread_db.so.1". > and then thrashes my hard drive for a very long time. I left it going for a > while (over half an hour, maybe closer to three quarters of an hour), and in > the end it seemed to have frozen the computer entirely. > > I only have about three quarters of a gigabyte of RAM, which may have > something to do with it... Huh. Can you then try ulimit -c unlimited, and upload the resulting core file somewhere? Mike
Bug#908449: firefox-esr_60.2.0esr-1~deb9u2.1_i386.deb not working
On 13/09/18 18:42, Mike Hommey wrote: Can you install the firefox-esr-dbgsym package from https://people.debian.org/~glandium/firefox-esr-dbgsym_60.2.0esr-1~deb9u2.1_i386.deb and run the following command: gdb firefox-esr At the gdb prompt, type the following commands: set pagination off run When I get to this point, it gives the output: Starting program: /usr/bin/firefox-esr [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib/i386-linux-gnu/libthread_db.so.1". and then thrashes my hard drive for a very long time. I left it going for a while (over half an hour, maybe closer to three quarters of an hour), and in the end it seemed to have frozen the computer entirely. I only have about three quarters of a gigabyte of RAM, which may have something to do with it... bt full x/i $eip And attach the output. Thanks Mike Cheers, Tim <><
Bug#908449: firefox-esr_60.2.0esr-1~deb9u2.1_i386.deb not working
On Thu, Sep 13, 2018 at 05:50:39PM +1200, Tim Makarios wrote: > Hi, > > Thanks for your work on this so far. I downloaded > firefox-esr_60.2.0esr-1~deb9u2.1_i386.deb, installed it via > > $ sudo apt-get install > > ./Downloads/firefox-esr_60.2.0esr-1~deb9u2.1_i386.deand tried to run it, > > but it crashes just as before. > > > $ cat /proc/cpuinfo > gives the following output: > > processor : 0 > > vendor_id : AuthenticAMD > > cpu family : 6 > > model : 8 > > model name : AMD Athlon(tm) XP 2400+ > > stepping: 1 > > cpu MHz : 1998.305 > > cache size : 256 KB > > physical id : 0 > > siblings: 1 > > core id : 0 > > cpu cores : 1 > > apicid : 0 > > initial apicid : 0 > > fdiv_bug: no > > f00f_bug: no > > coma_bug: no > > fpu : yes > > fpu_exception : yes > > cpuid level : 1 > > wp : yes > > flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge > > mca cmov pat pse36 mmx fxsr sse syscall mmxext 3dnowext 3dnow 3dnowprefetch > > vmmcall > > bugs: fxsave_leak sysret_ss_attrs spectre_v1 spectre_v2 > > spec_store_bypass > > bogomips: 3996.61 > > clflush size: 32 > > cache_alignment : 32 > > address sizes : 34 bits physical, 32 bits virtual > > power management: ts > > Let me know if you want more information. Can you install the firefox-esr-dbgsym package from https://people.debian.org/~glandium/firefox-esr-dbgsym_60.2.0esr-1~deb9u2.1_i386.deb and run the following command: gdb firefox-esr At the gdb prompt, type the following commands: set pagination off run bt full x/i $eip And attach the output. Thanks Mike
Bug#908449: firefox-esr_60.2.0esr-1~deb9u2.1_i386.deb not working
Hi, Thanks for your work on this so far. I downloaded firefox-esr_60.2.0esr-1~deb9u2.1_i386.deb, installed it via $ sudo apt-get install ./Downloads/firefox-esr_60.2.0esr-1~deb9u2.1_i386.deand tried to run it, but it crashes just as before. > $ cat /proc/cpuinfo gives the following output: processor : 0 vendor_id : AuthenticAMD cpu family : 6 model : 8 model name : AMD Athlon(tm) XP 2400+ stepping: 1 cpu MHz : 1998.305 cache size : 256 KB physical id : 0 siblings: 1 core id : 0 cpu cores : 1 apicid : 0 initial apicid : 0 fdiv_bug: no f00f_bug: no coma_bug: no fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 mmx fxsr sse syscall mmxext 3dnowext 3dnow 3dnowprefetch vmmcall bugs: fxsave_leak sysret_ss_attrs spectre_v1 spectre_v2 spec_store_bypass bogomips: 3996.61 clflush size: 32 cache_alignment : 32 address sizes : 34 bits physical, 32 bits virtual power management: ts Let me know if you want more information. Tim <><