Re: Alignment trap problem
[EMAIL PROTECTED] wrote: hi .. I am trying to run an application on davinci .. i have successfully cross compiled the source .. and while running on on board .. its giving alignment trap: bus error and application will abort.. I have tried the same source for x86 machine and it worked fine .. i know this has to do something with arm memory alignment access .. please can anybody tel me how to solve this alignment trap issue .. .. or can tel me where in source code i can fix this ?? arm can access 16bit words only on 2 byte aligned addresses and 32 byte words only on 4 byte aligned addresses. X86 does not care, so a lot of X86 code accesses 16 and 32 unaligned. You'll have to inspect your code and find the places where this happens, usual suspects are char* pointer which are casted to complex structs. The aligment trap should give you a hint where in your code this happened. http://netwinder.osuosl.org/users/b/brianbr/public_html/alignment.html http://www.mjmwired.net/kernel/Documentation/arm/mem_alignment Regards, Vladimir ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
RE: Alignment trap problem
Running your program with gdb can help you to find out where exactly the un-aligned memory access happens. Make sure program is compiled with debugging information. (gcc -g option) To run program with gdb on target, [EMAIL PROTECTED] ~]#gdb exe_name Then (gdb) run The program will start executing under control of gdb. When it crashes execute (gdb) backtrace. This should give you idea where exactly the alignment trap occurs. Regards, Kalpesh -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Vladimir Pantelic Sent: Friday, June 06, 2008 1:17 PM To: undisclosed-recipients: Cc: davinci-linux-open-source@linux.davincidsp.com Subject: Re: Alignment trap problem [EMAIL PROTECTED] wrote: hi .. I am trying to run an application on davinci .. i have successfully cross compiled the source .. and while running on on board .. its giving alignment trap: bus error and application will abort.. I have tried the same source for x86 machine and it worked fine .. i know this has to do something with arm memory alignment access .. please can anybody tel me how to solve this alignment trap issue .. .. or can tel me where in source code i can fix this ?? arm can access 16bit words only on 2 byte aligned addresses and 32 byte words only on 4 byte aligned addresses. X86 does not care, so a lot of X86 code accesses 16 and 32 unaligned. You'll have to inspect your code and find the places where this happens, usual suspects are char* pointer which are casted to complex structs. The aligment trap should give you a hint where in your code this happened. http://netwinder.osuosl.org/users/b/brianbr/public_html/alignment.html http://www.mjmwired.net/kernel/Documentation/arm/mem_alignment Regards, Vladimir ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Alignment trap problem
hi thank you for the reply this is what im getting Alignment trap: application(1198) PC=0x4074fcf0 Instr=0xe49c2004 Address=0x000e96d6 FSR 0x013 so how can i look in to this? [EMAIL PROTECTED] wrote: hi .. I am trying to run an application on davinci .. i have successfully cross compiled the source .. and while running on on board .. its giving alignment trap: bus error and application will abort.. I have tried the same source for x86 machine and it worked fine .. i know this has to do something with arm memory alignment access .. please can anybody tel me how to solve this alignment trap issue .. .. or can tel me where in source code i can fix this ?? arm can access 16bit words only on 2 byte aligned addresses and 32 byte words only on 4 byte aligned addresses. X86 does not care, so a lot of X86 code accesses 16 and 32 unaligned. You'll have to inspect your code and find the places where this happens, usual suspects are char* pointer which are casted to complex structs. The aligment trap should give you a hint where in your code this happened. http://netwinder.osuosl.org/users/b/brianbr/public_html/alignment.html http://www.mjmwired.net/kernel/Documentation/arm/mem_alignment Regards, Vladimir ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
RE: Alignment trap problem
gdb output .. it kils gdb itself so cant run backtrace oom-killer: gfp_mask=0x1d2 DMA per-cpu: cpu 0 hot: low 14, high 42, batch 7 cpu 0 cold: low 0, high 14, batch 7 Normal per-cpu: empty HighMem per-cpu: empty Free pages:1376kB (0kB HighMem) Active:24216 inactive:455 dirty:0 writeback:0 unstable:0 free:344 slab:886 mapped:23908 pagetables:113 DMA free:1376kB min:1400kB low:1748kB high:2100kB active:96864kB inactive:1820kB present:122880kB pages_scanned:100532 all_u nreclaimable? yes protections[]: 0 0 0 Normal free:0kB min:0kB low:0kB high:0kB active:0kB inactive:0kB present:0kB pages_scanned:0 all_unreclaimable? no protections[]: 0 0 0 HighMem free:0kB min:128kB low:160kB high:192kB active:0kB inactive:0kB present:0kB pages_scanned:0 all_unreclaimable? no protections[]: 0 0 0 DMA: 0*4kB 0*8kB 0*16kB 1*32kB 1*64kB 0*128kB 1*256kB 0*512kB 1*1024kB 0*2048kB 0*4096kB = 1376kB Normal: empty HighMem: empty Swap cache: add 0, delete 0, find 0/0, race 0+0 Out of Memory: Killed process 1246 (sshd). Running your program with gdb can help you to find out where exactly the un-aligned memory access happens. Make sure program is compiled with debugging information. (gcc -g option) To run program with gdb on target, [EMAIL PROTECTED] ~]#gdb exe_name Then (gdb) run The program will start executing under control of gdb. When it crashes execute (gdb) backtrace. This should give you idea where exactly the alignment trap occurs. Regards, Kalpesh -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Vladimir Pantelic Sent: Friday, June 06, 2008 1:17 PM To: undisclosed-recipients: Cc: davinci-linux-open-source@linux.davincidsp.com Subject: Re: Alignment trap problem [EMAIL PROTECTED] wrote: hi .. I am trying to run an application on davinci .. i have successfully cross compiled the source .. and while running on on board .. its giving alignment trap: bus error and application will abort.. I have tried the same source for x86 machine and it worked fine .. i know this has to do something with arm memory alignment access .. please can anybody tel me how to solve this alignment trap issue .. .. or can tel me where in source code i can fix this ?? arm can access 16bit words only on 2 byte aligned addresses and 32 byte words only on 4 byte aligned addresses. X86 does not care, so a lot of X86 code accesses 16 and 32 unaligned. You'll have to inspect your code and find the places where this happens, usual suspects are char* pointer which are casted to complex structs. The aligment trap should give you a hint where in your code this happened. http://netwinder.osuosl.org/users/b/brianbr/public_html/alignment.html http://www.mjmwired.net/kernel/Documentation/arm/mem_alignment Regards, Vladimir ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
RE: Alignment trap problem
If code is compiled with debugging information, gdb gives source level information. File name and line number when crash occurs. You can use -g option while compiling. There are other options as well. Please see gcc man page for that. Regards, Kalpesh -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Friday, June 06, 2008 2:35 PM To: davinci-linux-open-source@linux.davincidsp.com Subject: Re: Alignment trap problem hi thank you for the reply this is what im getting Alignment trap: application(1198) PC=0x4074fcf0 Instr=0xe49c2004 Address=0x000e96d6 FSR 0x013 so how can i look in to this? [EMAIL PROTECTED] wrote: hi .. I am trying to run an application on davinci .. i have successfully cross compiled the source .. and while running on on board .. its giving alignment trap: bus error and application will abort.. I have tried the same source for x86 machine and it worked fine .. i know this has to do something with arm memory alignment access .. please can anybody tel me how to solve this alignment trap issue .. .. or can tel me where in source code i can fix this ?? arm can access 16bit words only on 2 byte aligned addresses and 32 byte words only on 4 byte aligned addresses. X86 does not care, so a lot of X86 code accesses 16 and 32 unaligned. You'll have to inspect your code and find the places where this happens, usual suspects are char* pointer which are casted to complex structs. The aligment trap should give you a hint where in your code this happened. http://netwinder.osuosl.org/users/b/brianbr/public_html/alignment.html http://www.mjmwired.net/kernel/Documentation/arm/mem_alignment Regards, Vladimir ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
RE: Alignment trap problem
i know that .. its an application source code .. and its huge .. and debugging is enabled.. but gdb doesnt crash .. its giving following output oom-killer: gfp_mask=0x1d2 DMA per-cpu: cpu 0 hot: low 14, high 42, batch 7 cpu 0 cold: low 0, high 14, batch 7 Normal per-cpu: empty HighMem per-cpu: empty Free pages:1400kB (0kB HighMem) Active:24173 inactive:558 dirty:0 writeback:0 unstable:0 free:350 slab:877 mapped:23950 pagetables:101 DMA free:1400kB min:1400kB low:1748kB high:2100kB active:96692kB inactive:2232kB present:122880kB pages_scanned:100152 all_u nreclaimable? yes protections[]: 0 0 0 Normal free:0kB min:0kB low:0kB high:0kB active:0kB inactive:0kB present:0kB pages_scanned:0 all_unreclaimable? no protections[]: 0 0 0 HighMem free:0kB min:128kB low:160kB high:192kB active:0kB inactive:0kB present:0kB pages_scanned:0 all_unreclaimable? no protections[]: 0 0 0 DMA: 0*4kB 1*8kB 1*16kB 1*32kB 1*64kB 0*128kB 1*256kB 0*512kB 1*1024kB 0*2048kB 0*4096kB = 1400kB Normal: empty HighMem: empty Swap cache: add 0, delete 0, find 0/0, race 0+0 Out of Memory: Killed process 983 (rpc.statd). If code is compiled with debugging information, gdb gives source level information. File name and line number when crash occurs. You can use -g option while compiling. There are other options as well. Please see gcc man page for that. Regards, Kalpesh -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Friday, June 06, 2008 2:35 PM To: davinci-linux-open-source@linux.davincidsp.com Subject: Re: Alignment trap problem hi thank you for the reply this is what im getting Alignment trap: application(1198) PC=0x4074fcf0 Instr=0xe49c2004 Address=0x000e96d6 FSR 0x013 so how can i look in to this? [EMAIL PROTECTED] wrote: hi .. I am trying to run an application on davinci .. i have successfully cross compiled the source .. and while running on on board .. its giving alignment trap: bus error and application will abort.. I have tried the same source for x86 machine and it worked fine .. i know this has to do something with arm memory alignment access .. please can anybody tel me how to solve this alignment trap issue .. .. or can tel me where in source code i can fix this ?? arm can access 16bit words only on 2 byte aligned addresses and 32 byte words only on 4 byte aligned addresses. X86 does not care, so a lot of X86 code accesses 16 and 32 unaligned. You'll have to inspect your code and find the places where this happens, usual suspects are char* pointer which are casted to complex structs. The aligment trap should give you a hint where in your code this happened. http://netwinder.osuosl.org/users/b/brianbr/public_html/alignment.html http://www.mjmwired.net/kernel/Documentation/arm/mem_alignment Regards, Vladimir ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Alignment trap problem
[EMAIL PROTECTED] wrote: Out of Memory: Killed process 983 (rpc.statd). I think the last line is the most informative. Maybe add some (more) swap? -- Lisa ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Alignment trap problem
Lisa Parratt wrote: [EMAIL PROTECTED] wrote: Out of Memory: Killed process 983 (rpc.statd). I think the last line is the most informative. Maybe add some (more) swap? Actually, I think the program died due to memory exhaustion before ever running into the alignment trap. Running both gdb and the software seems to be too much. You should try post-mortem debugging with core dumps instead. regards, matthias -- * Matthias Welwarsky Project Manager Software Germany * This email and any files transmitted with it are confidential and are intended solely for the use of the individual or entity to which they are addressed. Access to this e-mail by anyone else is unauthorised. If you are not the intended recipient, any disclosure, copying, distribution or any action taken or omitted to be taken in reliance on it, is prohibited. E-mail messages are not necessarily secure. Archos does not accept responsibility for any changes made to this message after it was sent. ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Alignment trap problem
[EMAIL PROTECTED] wrote: this is what im getting Alignment trap: application(1198) PC=0x4074fcf0 Instr=0xe49c2004 Address=0x000e96d6 FSR 0x013 so how can i look in to this? Start gdb on the process but don't run it. Just type l *0x4074fcf0 and it should point you at the offending code (assuming you've got debug symbols compiled in). It won't give you a backtrace though - as Matthias mentioned, you'll need a coredump for that. This is also assuming that the code is not in a dynamically loaded library, and that /proc/sys/kernel/randomize_va_space is turned off (on newer kernels at least - it doesn't seem to be in the montavista one). HTH, Bernard. ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
