Re: Firefox trouble
On Tue, Jul 14, 2020 at 8:42 PM John Paul Adrian Glaubitz
<glaubitz@physik.fu-berlin.de> wrote:
>
> On 7/14/20 8:02 PM, Connor McLaughlan wrote:
> > Thread 1 "firefox" received signal SIGSEGV, Segmentation fault.
> > MOZ_CrashOOL (aLine=<optimized out>, aReason=0xfffff8010b632bf0
> > "js::jit::InitProcessExecutableMemory() failed")
> > at /build/firefox-YyDH69/firefox-62.0.3/mfbt/Assertions.cpp:33
> > 33 /build/firefox-YyDH69/firefox-62.0.3/mfbt/Assertions.cpp: No
> > such file or directory.
>
> See, that's the actual crash you were looking for.
>
> > (gdb) bt
> > #0 MOZ_CrashOOL (aLine=<optimized out>,
> > aReason=0xfffff8010b632bf0 "js::jit::InitProcessExecutableMemory() failed")
> > at /build/firefox-YyDH69/firefox-62.0.3/mfbt/Assertions.cpp:33
> > #1 0xfffff801078134e0 in NS_InitXPCOM2 (aResult=0x700001ea770,
> > aBinDirectory=<optimized out>,
> > aAppFileLocationProvider=<optimized out>)
> > at /build/firefox-YyDH69/firefox-62.0.3/xpcom/build/XPCOMInit.cpp:691
> > #2 0xfffff8010a365638 in ScopedXPCOMStartup::Initialize (this=0x700001ea770)
> > at /build/firefox-YyDH69/firefox-62.0.3/toolkit/xre/nsXREDirProvider.h:67
> > #3 0xfffff8010a36dce8 in XREMain::XRE_main (this=0x7feffffdcb8,
> > argc=<optimized out>, argv=<optimized out>,
> > aConfig=...) at
> > /build/firefox-YyDH69/firefox-62.0.3/toolkit/xre/nsAppRunner.cpp:4888
> > #4 0xfffff8010a36e1d0 in XRE_main (argc=<optimized out>,
> > argv=0x7fefffff378, aConfig=...)
> > at /build/firefox-YyDH69/firefox-62.0.3/toolkit/xre/nsAppRunner.cpp:4984
> > #5 0x0000010000006210 in do_main (argc=<optimized out>,
> > argv=0x7fefffff378, envp=0x7fefffff388)
> > at /build/firefox-YyDH69/firefox-62.0.3/build-browser/dist/include/mozilla/UniquePtr.h:326
> > #6 0x0000010000005a98 in main (argc=<optimized out>,
> > argv=0x7fefffff378, envp=0x7fefffff388)
> > at /build/firefox-YyDH69/firefox-62.0.3/browser/app/nsBrowserApp.cpp:311
> > (gdb)
>
> So, this is an intentional crash here:
>
> > /build/firefox-YyDH69/firefox-62.0.3/mfbt/Assertions.cpp:33
>
> which is triggered by the function "js::jit::InitProcessExecutableMemory()" failing.
>
> The Javascript memory management in Firefox can be nasty when running on an architecture
> with virtual address spaces larger than 47 bits [1]. I suggest using an earlier Firefox
> version (like 60).
>
> It might also be an unrelated bug for which NetBSD is carrying a fix.
>
> Adrian
>
> > [1] https://bugzilla.mozilla.org/show_bug.cgi?id=1441473
>
That is the problem: i am unable to locate any working version.
I would even use a firefox-52 or something if it was available.
On the repository there are:
firefox 62.0.3-1
firefox-esr 60.8.0esr-1
Both are segfaulting.
I searched snaphot the whole evening for available versions.
I could not find any sparc64 version of firefox 52.
The earliest i could find is firefox-esr 60.1.0esr-3, which is also segfaulting.
Any hints for any working firefox versions on debian sparc64 are welcome.
Regards,
Connor
Reply to: