Re: [e-users] e-crash issue
On Wed, 11 Jan 2023 18:22:56 +0900 Masaru Nomiya said: > Hello, > > In the Message; > > Subject : Re: [e-users] e-crash issue > Message-ID : <20230111083906.22c9ccc18ce34767e012f...@rasterman.com> > Date & Time: Wed, 11 Jan 2023 08:39:06 + > > [CH] == Carsten Haitzler has written: > > CH> On Wed, 11 Jan 2023 11:00:48 +0900 Masaru Nomiya > CH> said: > > [...] > MN> > I solidified the e-log.log into a log.tgz and uploaded it to the > MN> > following location; > MN> > > MN> > > https://drive.google.com/file/d/100t8RwoGCtFdqAvZmWKpuVPA1UH0qca5/view?usp=sharing > > CH> I see like complete eina log output (not useful). you've raised your > CH> eina log level to something really extreme... but there is nothing in > CH> there from asan. so .. you didn't get that log, or asan is not working. i > CH> suggest you reduce your eina log levels again to normal > > I got the .e-log.log under this setting; > > export EINA_LOG_LEVEL=4 don't set this or set it to 1. we don't need/want those logs. we want the asan log/output. > Is this level too high? > > What setting is "normal level"? > > Or is this a bad build condition? > > export > ASAN_OPTIONS=detect_stack_use_after_return=0:detect_odr_violation=0:detect_leaks=0:abort_on_error=1:new_delete_type_mismatch > =0 this is about right. > Regards. > > --- > ┏━━┓彡 野宮 賢 mail-to: m.nomiya+enlight @ gmail.com > ┃\/彡 > ┗━━┛ "No Windows, no gains!" ... "Why, I am wrong?" > -- Bill -- > > > ___ > enlightenment-users mailing list > enlightenment-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/enlightenment-users > -- - Codito, ergo sum - "I code, therefore I am" -- Carsten Haitzler - ras...@rasterman.com ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
Re: [e-users] e-crash issue
Hello, In the Message; Subject: Re: [e-users] e-crash issue Message-ID : <20230111083906.22c9ccc18ce34767e012f...@rasterman.com> Date & Time: Wed, 11 Jan 2023 08:39:06 + [CH] == Carsten Haitzler has written: CH> On Wed, 11 Jan 2023 11:00:48 +0900 Masaru Nomiya said: [...] MN> > I solidified the e-log.log into a log.tgz and uploaded it to the MN> > following location; MN> > MN> > https://drive.google.com/file/d/100t8RwoGCtFdqAvZmWKpuVPA1UH0qca5/view?usp=sharing CH> I see like complete eina log output (not useful). you've raised your eina log CH> level to something really extreme... but there is nothing in there from asan. CH> so .. you didn't get that log, or asan is not working. i suggest you reduce CH> your eina log levels again to normal I got the .e-log.log under this setting; export EINA_LOG_LEVEL=4 Is this level too high? What setting is "normal level"? Or is this a bad build condition? export ASAN_OPTIONS=detect_stack_use_after_return=0:detect_odr_violation=0:detect_leaks=0:abort_on_error=1:new_delete_type_mismatch =0 Regards. --- ┏━━┓彡 野宮 賢 mail-to: m.nomiya+enlight @ gmail.com ┃\/彡 ┗━━┛ "No Windows, no gains!" ... "Why, I am wrong?" -- Bill -- ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
Re: [e-users] e-crash issue
On Wed, 11 Jan 2023 11:00:48 +0900 Masaru Nomiya said: > Hello, > > In the Message; > > Subject : Re: [e-users] e-crash issue > Message-ID : <20221221105449.7e143081d6f53eac37983...@rasterman.com> > Date & Time: Wed, 21 Dec 2022 10:54:49 + > > [CH] == Carsten Haitzler has written: > > CH> On Wed, 21 Dec 2022 19:02:45 +0900 Masaru Nomiya > CH> said: > > [...] > MN> > Anyway, I uploaded it here under the file nmae dump.tgz; > MN> > > MN> > > https://drive.google.com/file/d/1VLvws1FYbjOUsAw3z8RLDSCzy9QQZjhl/view?usp=sharing > > CH> i need ~/.e-log.log to see asan's output... th crashdump is also > CH> useful-but the asan log in ~/.e-log.log is very important. here > CH> what i see is a crash inside malloc itself, so i think by now > CH> everything will have gone bad and memory malloc relies on > CH> tracking allocations is somehow messed up. > > CH> you might have to go back to good old valgrind... > > Finally, I was able to get .e-log.log. > > No matter how many times I did the debug build, the .e-log.log was not > generated, and I finally figured out that it was caused by gcc13 > installed with the Tumbleweed update, and added the following; > ASAN_OPTION > > detect_stack_use_after_return=0 that is actually a useful thing in asan it should trap and find and report. https://www.enlightenment.org/contrib/efl-debug has an asan section and suggests some asan options that won't impact e/efl. > to ASAN_OPTION was the solution. > > Anyway, today, I got an e-crash. > > When I downloaded the 02_o.png file from this site; > > https://internet.watch.impress.co.jp/img/iw/docs/1469/183/html/02_o.png.html > > e-crash appeared. > > I solidified the e-log.log into a log.tgz and uploaded it to the > following location; > > https://drive.google.com/file/d/100t8RwoGCtFdqAvZmWKpuVPA1UH0qca5/view?usp=sharing I see like complete eina log output (not useful). you've raised your eina log level to something really extreme... but there is nothing in there from asan. so .. you didn't get that log, or asan is not working. i suggest you reduce your eina log levels again to normal > Thanks in advance. > > --- > ┏━━┓彡 野宮 賢 mail-to: m.nomiya+enlight @ gmail.com > ┃\/彡 > ┗━━┛ "A bachelor’s degree still holds prestige as a ticket to the > middle class, but its value has received increasing scrutiny. > In the last several years, rising tuition and student loan debt > have led more Americans to reconsider an investment in > postsecondary education." > > -- Washington Post -- > > > ___ > enlightenment-users mailing list > enlightenment-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/enlightenment-users > -- - Codito, ergo sum - "I code, therefore I am" -- Carsten Haitzler - ras...@rasterman.com ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
Re: [e-users] e-crash issue
Hello, In the Message; Subject: Re: [e-users] e-crash issue Message-ID : <20221221105449.7e143081d6f53eac37983...@rasterman.com> Date & Time: Wed, 21 Dec 2022 10:54:49 + [CH] == Carsten Haitzler has written: CH> On Wed, 21 Dec 2022 19:02:45 +0900 Masaru Nomiya said: [...] MN> > Anyway, I uploaded it here under the file nmae dump.tgz; MN> > MN> > https://drive.google.com/file/d/1VLvws1FYbjOUsAw3z8RLDSCzy9QQZjhl/view?usp=sharing CH> i need ~/.e-log.log to see asan's output... th crashdump is also CH> useful-but the asan log in ~/.e-log.log is very important. here CH> what i see is a crash inside malloc itself, so i think by now CH> everything will have gone bad and memory malloc relies on CH> tracking allocations is somehow messed up. CH> you might have to go back to good old valgrind... Finally, I was able to get .e-log.log. No matter how many times I did the debug build, the .e-log.log was not generated, and I finally figured out that it was caused by gcc13 installed with the Tumbleweed update, and added the following; ASAN_OPTION detect_stack_use_after_return=0 to ASAN_OPTION was the solution. Anyway, today, I got an e-crash. When I downloaded the 02_o.png file from this site; https://internet.watch.impress.co.jp/img/iw/docs/1469/183/html/02_o.png.html e-crash appeared. I solidified the e-log.log into a log.tgz and uploaded it to the following location; https://drive.google.com/file/d/100t8RwoGCtFdqAvZmWKpuVPA1UH0qca5/view?usp=sharing Thanks in advance. --- ┏━━┓彡 野宮 賢 mail-to: m.nomiya+enlight @ gmail.com ┃\/彡 ┗━━┛ "A bachelor’s degree still holds prestige as a ticket to the middle class, but its value has received increasing scrutiny. In the last several years, rising tuition and student loan debt have led more Americans to reconsider an investment in postsecondary education." -- Washington Post -- ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
Re: [e-users] e-crash issue
On Wed, 21 Dec 2022 19:02:45 +0900 Masaru Nomiya said: > Hello, > > In the Message; > > Subject : Re: [e-users] e-crash issue > Message-ID : <20221216054249.15580dfc6752064485701...@rasterman.com> > Date & Time: Fri, 16 Dec 2022 05:42:49 + > > [CH] == Carsten Haitzler has written: > > CH> On Thu, 15 Dec 2022 18:09:13 +0900 Masaru Nomiya > CH> said: > > [...] > CH> you can read their PKGBUILD files to see how > CH> efl and e are built with asan. > > CH> https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=efl-git-asan > CH> > https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=enlightenment-git-asan > > Sorry , but I had done it,before. m(_ _)m > > BTW, I did so, and I got a crash dump just now, > But, files's name is .e-crashdump.txt (before, the file name was > indeed .e-log.log). > > Anyway, I uploaded it here under the file nmae dump.tgz; > > > https://drive.google.com/file/d/1VLvws1FYbjOUsAw3z8RLDSCzy9QQZjhl/view?usp=sharing i need ~/.e-log.log to see asan's output... th crashdump is also useful-but the asan log in ~/.e-log.log is very important. here what i see is a crash inside malloc itself, so i think by now everything will have gone bad and memory malloc relies on tracking allocations is somehow messed up. you might have to go back to good old valgrind... > > Regards, > > --- > ┏━━┓彡 野宮 賢 mail-to: nomiya @ galaxy.dti.ne.jp > ┃\/彡 > ┗━━┛ "Bill! > You married with Computer. >Not with Me!" >"No..., with money." > > > ___ > enlightenment-users mailing list > enlightenment-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/enlightenment-users > -- - Codito, ergo sum - "I code, therefore I am" -- Carsten Haitzler - ras...@rasterman.com ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
Re: [e-users] e-crash issue
Hello, In the Message; Subject: Re: [e-users] e-crash issue Message-ID : <20221216054249.15580dfc6752064485701...@rasterman.com> Date & Time: Fri, 16 Dec 2022 05:42:49 + [CH] == Carsten Haitzler has written: CH> On Thu, 15 Dec 2022 18:09:13 +0900 Masaru Nomiya said: [...] CH> you can read their PKGBUILD files to see how CH> efl and e are built with asan. CH> https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=efl-git-asan CH> https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=enlightenment-git-asan Sorry , but I had done it,before. m(_ _)m BTW, I did so, and I got a crash dump just now, But, files's name is .e-crashdump.txt (before, the file name was indeed .e-log.log). Anyway, I uploaded it here under the file nmae dump.tgz; https://drive.google.com/file/d/1VLvws1FYbjOUsAw3z8RLDSCzy9QQZjhl/view?usp=sharing Regards, --- ┏━━┓彡 野宮 賢 mail-to: nomiya @ galaxy.dti.ne.jp ┃\/彡 ┗━━┛ "Bill! You married with Computer. Not with Me!" "No..., with money." ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
Re: [e-users] e-crash issue
On Thu, 15 Dec 2022 18:09:13 +0900 Masaru Nomiya said: you have no symnbols. no gdb debug. i see 3 crashes in that log. (chrashdump keeps being appended to each crash): Thread 1 (Thread 0x7f6dae712e00 (LWP 17060) "enlightenment"): #0 0x7f6dafa83bf2 in pause () at /lib64/libc.so.6 #1 0x7f6daf9e2460 in () at /lib64/libc.so.6 #2 0x563c0a64f928 in e_notification_notify_close () Thread 1 (Thread 0x7f4b0e0c0e00 (LWP 23723) "enlightenment"): #0 0x7f4b0f483bf2 in pause () at /lib64/libc.so.6 #1 0x7f4b0f3e2460 in () at /lib64/libc.so.6 #2 0x5628ee3b4928 in e_notification_notify_close () Thread 1 (Thread 0x7f5f809bfe00 (LWP 2717) "enlightenment"): #0 0x7f5f81d81bf2 in pause () at /lib64/libc.so.6 #1 0x7f5f81ce0460 in () at /lib64/libc.so.6 #2 0x7f5f73702090 in () #at /usr/lib64/enlightenment/modules/mixer/linux-gnu-x86_64-dev-0.25.99/module.so each different. the first 2 i assume are crashes in libc - probably due to passing in a bad ptr to a str func or a malloc related bug (double-free. free junk pointer etc.). the last - who knows. access of some incorrect pointer. you have no debug info to use to do anything. i suggest you read up on asan (see enlightenment.org/contrib/efl-debug ) and rebuild efl and e with asan support (and gdb debug of course) then submit the asan logs (from ~/.e-log.log). arch linux has efl-git-asan and enlightenment-git-asan aur packages just for this purpose. you can read their PKGBUILD files to see how efl and e are built with asan. https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=efl-git-asan https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=enlightenment-git-asan (PKGBUILD files are literally shell scripts pretty much...). > Hello, > > The flashing phenomenon occurred once, but I was unable to record it > due to a concurrent e-crash, and will do so at a later date . > > Another problem now. > > I often download web files using the browser Vivaldi, and occasionally > (but I feel it is more often than not) I get an e-crash. > Until now, I thought it was probably caused by Vivaldi, butToday, > after updating pipewire and wireplumber. > >$ systemctl --user daemon-reload >$ systemctl --user pipewire pipewire-pulse wireplumber > > on the shelf, I couldn't see the Mixer icon on the shelf anymore, so I > clicked on that part and it caused e-crash, and I panicked and pressed > the F1 key. > > This made me wonder if the e-crash could have been caused by something > else. > So, I uploaded the .e-crahdump.txt file. > > I have put the file as e-crashdump.txt.gz at the following location. > > https://drive.google.com/file/d/1asT2-k8-ykkKWBEAQkjoDnJIKN-W4SHQ/view?usp=share_link > > Could you have look? > > I would appreciate your advice. > > --- > ┏━━┓彡 野宮 賢 mail-to: nomiya @ galaxy.dti.ne.jp > ┃\/彡 > ┗━━┛ "A bachelor’s degree still holds prestige as a ticket to the > middle class, but its value has received increasing scrutiny. > In the last several years, rising tuition and student loan debt > have led more Americans to reconsider an investment in > postsecondary education." > > -- Washington Post -- > > > ___ > enlightenment-users mailing list > enlightenment-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/enlightenment-users > -- - Codito, ergo sum - "I code, therefore I am" -- Carsten Haitzler - ras...@rasterman.com ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
[e-users] e-crash issue
Hello, The flashing phenomenon occurred once, but I was unable to record it due to a concurrent e-crash, and will do so at a later date . Another problem now. I often download web files using the browser Vivaldi, and occasionally (but I feel it is more often than not) I get an e-crash. Until now, I thought it was probably caused by Vivaldi, butToday, after updating pipewire and wireplumber. $ systemctl --user daemon-reload $ systemctl --user pipewire pipewire-pulse wireplumber on the shelf, I couldn't see the Mixer icon on the shelf anymore, so I clicked on that part and it caused e-crash, and I panicked and pressed the F1 key. This made me wonder if the e-crash could have been caused by something else. So, I uploaded the .e-crahdump.txt file. I have put the file as e-crashdump.txt.gz at the following location. https://drive.google.com/file/d/1asT2-k8-ykkKWBEAQkjoDnJIKN-W4SHQ/view?usp=share_link Could you have look? I would appreciate your advice. --- ┏━━┓彡 野宮 賢 mail-to: nomiya @ galaxy.dti.ne.jp ┃\/彡 ┗━━┛ "A bachelor’s degree still holds prestige as a ticket to the middle class, but its value has received increasing scrutiny. In the last several years, rising tuition and student loan debt have led more Americans to reconsider an investment in postsecondary education." -- Washington Post -- ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users