|
From: Ilya S. <ily...@gm...> - 2012-10-31 17:49:20
|
Dear Developers, I just trying to use valgrind on the target with description in topic and got an error on the running valgrind even without arguments. Please, if it's possible, gimme some support. / # ulimit -a time(seconds) unlimited file(blocks) unlimited data(kb) unlimited stack(kb) 512 coredump(blocks) 0 memory(kb) unlimited locked memory(kb) 64 process 950 nofiles 1024 vmemory(kb) unlimited locks unlimited / # valgrind -v -v -v -d -d -d --trace-syscalls=yes --3518:1:debuglog DebugLog system started by Stage 1, level 3 logging requested --3518:1:launcher no tool requested, defaulting to 'memcheck' --3518:1:launcher no client specified, defaulting platform to 'mips32-linux' --3518:1:launcher launching /home/ilya/valgrind_mips/lib/valgrind/memcheck-mips32-linux --3518:1:debuglog DebugLog system started by Stage 2 (main), level 3 logging requested --3518:1:main Welcome to Valgrind version 3.9.0.SVN debug logging --3518:1:main Checking current stack is plausible --3518:1:main Checking initial stack was noted --3518:1:main Starting the address space manager --3518:2:aspacem sp_at_startup = 0x007fbc7630 (supplied) --3518:2:aspacem minAddr = 0x0004000000 (computed) --3518:2:aspacem maxAddr = 0x007fbc6fff (computed) --3518:2:aspacem cStart = 0x0004000000 (computed) --3518:2:aspacem vStart = 0x0041de4000 (computed) --3518:2:aspacem suggested_clstack_top = 0x007ebc7fff (computed) --3518:2:aspacem <<< SHOW_SEGMENTS: Initial layout (5 segments, 0 segnames) --3518:2:aspacem 0: RSVN 0000000000-0003ffffff 64m ----- SmFixed --3518:2:aspacem 1: 0004000000-0041de3fff 989m --3518:2:aspacem 2: RSVN 0041de4000-0041de4fff 4096 ----- SmFixed --3518:2:aspacem 3: 0041de5000-007fbc6fff 989m --3518:2:aspacem 4: RSVN 007fbc7000-00ffffffff 2052m ----- SmFixed --3518:2:aspacem >>> --3518:2:aspacem Reading /proc/self/maps --3518:2:aspacem <<< SHOW_SEGMENTS: With contents of /proc/self/maps (15 segments, 1 segnames) --3518:2:aspacem ( 0) /disk/lib/valgrind/memcheck-mips32-linux --3518:2:aspacem 0: RSVN 0000000000-00003fffff 4194304 ----- SmFixed --3518:2:aspacem 1: FILE 0000400000-0000403fff 16384 r---- d=0x803 i=13419112 o=0 (0) --3518:2:aspacem 2: RSVN 0000404000-0003ffffff 59m ----- SmFixed --3518:2:aspacem 3: 0004000000-0037ffffff 832m --3518:2:aspacem 4: FILE 0038000000-0038397fff 3768320 r-x-- d=0x803 i=13419112 o=65536 (0) --3518:2:aspacem 5: 0038398000-00383a3fff 49152 --3518:2:aspacem 6: FILE 00383a4000-00383b3fff 65536 rw--- d=0x803 i=13419112 o=3817472 (0) --3518:2:aspacem 7: ANON 00383b4000-0038e7bfff 10m rwx-- --3518:2:aspacem 8: 0038e7c000-0041de3fff 143m --3518:2:aspacem 9: RSVN 0041de4000-0041de4fff 4096 ----- SmFixed --3518:2:aspacem 10: 0041de5000-007fba3fff 989m --3518:2:aspacem 11: ANON 007fba4000-007fbc7fff 147456 rwx-- --3518:2:aspacem 12: RSVN 007fbc8000-007fff3fff 4374528 ----- SmFixed --3518:2:aspacem 13: ANON 007fff4000-007fff7fff 16384 r-x-- --3518:2:aspacem 14: RSVN 007fff8000-00ffffffff 2048m ----- SmFixed --3518:2:aspacem >>> --3518:1:main Address space manager is running --3518:1:main Starting the dynamic memory manager --3518:0:aspacem <<< SHOW_SEGMENTS: out_of_memory (15 segments, 1 segnames) --3518:0:aspacem ( 0) /disk/lib/valgrind/memcheck-mips32-linux --3518:0:aspacem 0: RSVN 0000000000-00003fffff 4194304 ----- SmFixed --3518:0:aspacem 1: FILE 0000400000-0000403fff 16384 r---- d=0x803 i=13419112 o=0 (0) --3518:0:aspacem 2: RSVN 0000404000-0003ffffff 59m ----- SmFixed --3518:0:aspacem 3: 0004000000-0037ffffff 832m --3518:0:aspacem 4: FILE 0038000000-0038397fff 3768320 r-x-- d=0x803 i=13419112 o=65536 (0) --3518:0:aspacem 5: 0038398000-00383a3fff 49152 --3518:0:aspacem 6: FILE 00383a4000-00383b3fff 65536 rw--- d=0x803 i=13419112 o=3817472 (0) --3518:0:aspacem 7: ANON 00383b4000-0038e7bfff 10m rwx-- --3518:0:aspacem 8: 0038e7c000-0041de3fff 143m --3518:0:aspacem 9: RSVN 0041de4000-0041de4fff 4096 ----- SmFixed --3518:0:aspacem 10: 0041de5000-007fba3fff 989m --3518:0:aspacem 11: ANON 007fba4000-007fbc7fff 147456 rwx-- --3518:0:aspacem 12: RSVN 007fbc8000-007fff3fff 4374528 ----- SmFixed --3518:0:aspacem 13: ANON 007fff4000-007fff7fff 16384 r-x-- --3518:0:aspacem 14: RSVN 007fff8000-00ffffffff 2048m ----- SmFixed --3518:0:aspacem >>> --3518-- core : 0/ 0 max/curr mmap'd, 0/0 unsplit/split sb unmmap'd, 0/ 0 max/curr, 0/ 0 totalloc-blocks/bytes, 0 searches 4 rzB --3518-- tool : 0/ 0 max/curr mmap'd, 0/0 unsplit/split sb unmmap'd, 0/ 0 max/curr, 0/ 0 totalloc-blocks/bytes, 0 searches 4 rzB --3518-- dinfo : 0/ 0 max/curr mmap'd, 0/0 unsplit/split sb unmmap'd, 0/ 0 max/curr, 0/ 0 totalloc-blocks/bytes, 0 searches 4 rzB --3518-- (null) : 0/ 0 max/curr mmap'd, 0/0 unsplit/split sb unmmap'd, 0/ 0 max/curr, 0/ 0 totalloc-blocks/bytes, 0 searches 0 rzB --3518-- demangle: 0/ 0 max/curr mmap'd, 0/0 unsplit/split sb unmmap'd, 0/ 0 max/curr, 0/ 0 totalloc-blocks/bytes, 0 searches 4 rzB --3518-- exectxt : 0/ 0 max/curr mmap'd, 0/0 unsplit/split sb unmmap'd, 0/ 0 max/curr, 0/ 0 totalloc-blocks/bytes, 0 searches 4 rzB --3518-- errors : 0/ 0 max/curr mmap'd, 0/0 unsplit/split sb unmmap'd, 0/ 0 max/curr, 0/ 0 totalloc-blocks/bytes, 0 searches 4 rzB --3518-- ttaux : 0/ 0 max/curr mmap'd, 0/0 unsplit/split sb unmmap'd, 0/ 0 max/curr, 0/ 0 totalloc-blocks/bytes, 0 searches 4 rzB ==3518== ==3518== Valgrind's memory management: out of memory: ==3518== newSuperblock's request for 4194304 bytes failed. ==3518== 11468800 bytes have already been allocated. ==3518== Valgrind cannot continue. Sorry. ==3518== ==3518== There are several possible reasons for this. ==3518== - You have some kind of memory limit in place. Look at the ==3518== output of 'ulimit -a'. Is there a limit on the size of ==3518== virtual memory or address space? ==3518== - You have run out of swap space. ==3518== - Valgrind has a bug. If you think this is the case or you are ==3518== not sure, please let us know and we'll try to fix it. ==3518== Please note that programs can take substantially more memory than ==3518== normal when running under Valgrind tools, eg. up to twice or ==3518== more, depending on the tool. On a 64-bit machine, Valgrind ==3518== should be able to make use of up 32GB memory. On a 32-bit ==3518== machine, Valgrind should be able to use all the memory available ==3518== to a single process, up to 4GB if that's how you have your ==3518== kernel configured. Most 32-bit Linux setups allow a maximum of ==3518== 3GB per process. ==3518== ==3518== Whatever the reason, Valgrind cannot continue. Sorry. $ cat config.log | head This file contains any messages produced by compilers while running configure, to aid debugging if configure makes a mistake. It was created by Valgrind configure 3.9.0.SVN, which was generated by GNU Autoconf 2.68. Invocation command line was $ ./configure --host=mipsel-linux-gnu CC=mipsisa32r2el-timesys-linux-gnu-gcc --prefix=/home/ilya/valgrind_mips/ --with-page-size=16 |
|
From: Philippe W. <phi...@sk...> - 2012-10-31 18:32:11
|
On Thu, 2012-11-01 at 00:48 +0700, Ilya Smelykh wrote: > > --3518:1:main Address space manager is running > --3518:1:main Starting the dynamic memory manager if you run: strace -f -s 50 -o v.out valgrind -v -v -v -d -d -d --trace-syscalls=yes then v.out should contain a syscall to write the "Starting the dynamic memory manager. 22888 write(2, "--22888:1:main Starting the dynamic memory man"..., 55) = 55 If all goes well, a successful mmap syscall should follow: 22888 old_mmap(0x41c70000, 4194304, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, 0, 0) = 0x41c70000 If you have a failing syscall, then the errno might give an idea about what is wrong. Otherwise, something might not be ok in the way the valgrind aspacemgr finds a "free" zone in the list of segments. Philippe |
|
From: John R. <jr...@bi...> - 2012-10-31 18:48:42
|
The ending address 0041de4fff is not the end of a 16KiB page: --3518:2:aspacem 9: RSVN 0041de4000-0041de4fff 4096 ----- SmFixed --3518:2:aspacem 10: 0041de5000-007fba3fff 989m Neither is the _next_ address 0041de5000 the beginning of a 16KiB page. There was a similar problem on ARM, now fixed: https://bugs.kde.org/show_bug.cgi?id=290974 Philippe's debugging suggestion is good: use strace, look for parameters and errno of a failing call to mmap or mmap2. -- |
|
From: Philippe W. <phi...@sk...> - 2012-10-31 19:43:47
|
On Wed, 2012-10-31 at 11:49 -0700, John Reiser wrote: > The ending address 0041de4fff is not the end of a 16KiB page: > --3518:2:aspacem 9: RSVN 0041de4000-0041de4fff 4096 ----- SmFixed > --3518:2:aspacem 10: 0041de5000-007fba3fff 989m > Neither is the _next_ address 0041de5000 the beginning of a 16KiB page. Interesting observation. It looks like the /proc/self/maps is already not aligned on 16Kb pages. Is that normal if the kernel is configured to be 16Kb pages ? What does getconf -a | grep PAGESIZE give ? on gcc49, it gives PAGESIZE 16384 and there, the /proc/self/maps reading is properly page aligned. Maybe the kernel is not configured for 16Kb pages ? Otherwise, strace will help to see what is going on for failing mmaps. Philippe |
|
From: Ilya S. <ily...@gm...> - 2012-11-01 03:54:31
|
Kernel configured to 16K page size, there is no getconf tool, but getpagesize() returns 16384. I olso checked valgrind with strace: old_mmap(0x41db7000, 4194304, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, 0, 0) = -1 EINVAL (Invalid argument) Looks like address is not 16K aligned. Ilya. 2012/11/1 Philippe Waroquiers <phi...@sk...>: > On Wed, 2012-10-31 at 11:49 -0700, John Reiser wrote: >> The ending address 0041de4fff is not the end of a 16KiB page: >> --3518:2:aspacem 9: RSVN 0041de4000-0041de4fff 4096 ----- SmFixed >> --3518:2:aspacem 10: 0041de5000-007fba3fff 989m >> Neither is the _next_ address 0041de5000 the beginning of a 16KiB page. > > Interesting observation. > > It looks like the /proc/self/maps is already not aligned on 16Kb > pages. Is that normal if the kernel is configured to be 16Kb pages ? > What does > getconf -a | grep PAGESIZE > give ? > on gcc49, it gives > PAGESIZE 16384 > and there, the /proc/self/maps reading is properly page aligned. > > Maybe the kernel is not configured for 16Kb pages ? > > Otherwise, strace will help to see what is going on for failing > mmaps. > > Philippe > > > > ------------------------------------------------------------------------------ > Everyone hates slow websites. So do we. > Make your web apps faster with AppDynamics > Download AppDynamics Lite for free today: > http://p.sf.net/sfu/appdyn_sfd2d_oct > _______________________________________________ > Valgrind-developers mailing list > Val...@li... > https://lists.sourceforge.net/lists/listinfo/valgrind-developers |
|
From: Dejan J. <dej...@rt...> - 2012-11-06 10:43:27
|
From your previous mail: > --3518:1:main Starting the address space manager > --3518:2:aspacem sp_at_startup = 0x007fbc7630 (supplied) > --3518:2:aspacem minAddr = 0x0004000000 (computed) > --3518:2:aspacem maxAddr = 0x007fbc6fff (computed) > --3518:2:aspacem cStart = 0x0004000000 (computed) > --3518:2:aspacem vStart = 0x0041de4000 (computed) > --3518:2:aspacem suggested_clstack_top = 0x007ebc7fff (computed) > --3518:2:aspacem <<< SHOW_SEGMENTS: Initial layout (5 segments, 0 segnames) > --3518:2:aspacem 0: RSVN 0000000000-0003ffffff 64m ----- SmFixed > --3518:2:aspacem 1: 0004000000-0041de3fff 989m > --3518:2:aspacem 2: RSVN 0041de4000-0041de4fff 4096 ----- SmFixed > --3518:2:aspacem 3: 0041de5000-007fbc6fff 989m > --3518:2:aspacem 4: RSVN 007fbc7000-00ffffffff 2052m ----- SmFixed I think that your configuration is not good because: > --3518:2:aspacem 2: RSVN 0041de4000-0041de4fff 4096 ----- SmFixed 4096 is VKI_PAGE_SIZE and that is your PAGE_SIZE (see coregrind/m_aspacemgr/aspacemgr-linux.c) Can you see your config.h file and send me values for MIPS_PAGE_SHIFT and SHM_ALIGNMENT. Regards, Dejan On 11/01/2012 04:54 AM, Ilya Smelykh wrote: > Kernel configured to 16K page size, there is no getconf tool, but > getpagesize() returns 16384. > > I olso checked valgrind with strace: > old_mmap(0x41db7000, 4194304, PROT_READ|PROT_WRITE|PROT_EXEC, > MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, 0, 0) = -1 EINVAL (Invalid > argument) > > Looks like address is not 16K aligned. > > Ilya. > > 2012/11/1 Philippe Waroquiers <phi...@sk...>: >> On Wed, 2012-10-31 at 11:49 -0700, John Reiser wrote: >>> The ending address 0041de4fff is not the end of a 16KiB page: >>> --3518:2:aspacem 9: RSVN 0041de4000-0041de4fff 4096 ----- SmFixed >>> --3518:2:aspacem 10: 0041de5000-007fba3fff 989m >>> Neither is the _next_ address 0041de5000 the beginning of a 16KiB page. >> Interesting observation. >> >> It looks like the /proc/self/maps is already not aligned on 16Kb >> pages. Is that normal if the kernel is configured to be 16Kb pages ? >> What does >> getconf -a | grep PAGESIZE >> give ? >> on gcc49, it gives >> PAGESIZE 16384 >> and there, the /proc/self/maps reading is properly page aligned. >> >> Maybe the kernel is not configured for 16Kb pages ? >> >> Otherwise, strace will help to see what is going on for failing >> mmaps. >> >> Philippe >> >> >> >> ------------------------------------------------------------------------------ >> Everyone hates slow websites. So do we. >> Make your web apps faster with AppDynamics >> Download AppDynamics Lite for free today: >> http://p.sf.net/sfu/appdyn_sfd2d_oct >> _______________________________________________ >> Valgrind-developers mailing list >> Val...@li... >> https://lists.sourceforge.net/lists/listinfo/valgrind-developers > ------------------------------------------------------------------------------ > Everyone hates slow websites. So do we. > Make your web apps faster with AppDynamics > Download AppDynamics Lite for free today: > http://p.sf.net/sfu/appdyn_sfd2d_oct > _______________________________________________ > Valgrind-developers mailing list > Val...@li... > https://lists.sourceforge.net/lists/listinfo/valgrind-developers > |
|
From: Ilya S. <ily...@gm...> - 2012-11-06 10:26:13
|
/* configured default page size 4k */ #define MIPS_PAGE_SHIFT 14 /* configured memory alignment 2*PAGE_SIZE */ #define SHM_ALIGNMENT 2*(1UL << MIPS_PAGE_SHIFT) 2012/11/6 Dejan Jevtic <dej...@rt...> > From your previous mail: > > > --3518:1:main Starting the address space manager > > --3518:2:aspacem sp_at_startup = 0x007fbc7630 (supplied) > > --3518:2:aspacem minAddr = 0x0004000000 (computed) > > --3518:2:aspacem maxAddr = 0x007fbc6fff (computed) > > --3518:2:aspacem cStart = 0x0004000000 (computed) > > --3518:2:aspacem vStart = 0x0041de4000 (computed) > > --3518:2:aspacem suggested_clstack_top = 0x007ebc7fff (computed) > > --3518:2:aspacem <<< SHOW_SEGMENTS: Initial layout (5 segments, 0 > segnames) > > --3518:2:aspacem 0: RSVN 0000000000-0003ffffff 64m ----- SmFixed > > --3518:2:aspacem 1: 0004000000-0041de3fff 989m > > --3518:2:aspacem 2: RSVN 0041de4000-0041de4fff 4096 ----- SmFixed > > --3518:2:aspacem 3: 0041de5000-007fbc6fff 989m > > --3518:2:aspacem 4: RSVN 007fbc7000-00ffffffff 2052m ----- SmFixed > > I think that your configuration is not good because: > > > --3518:2:aspacem 2: RSVN 0041de4000-0041de4fff 4096 ----- SmFixed > > 4096 is VKI_PAGE_SIZE and that is your PAGE_SIZE (see > coregrind/m_aspacemgr/**aspacemgr-linux.c) > > Can you see your config.h file and send me values for MIPS_PAGE_SHIFT and > SHM_ALIGNMENT. > > Regards, > Dejan > > On 11/01/2012 04:54 AM, Ilya Smelykh wrote: > >> Kernel configured to 16K page size, there is no getconf tool, but >> getpagesize() returns 16384. >> >> I olso checked valgrind with strace: >> old_mmap(0x41db7000, 4194304, PROT_READ|PROT_WRITE|PROT_**EXEC, >> MAP_PRIVATE|MAP_FIXED|MAP_**ANONYMOUS, 0, 0) = -1 EINVAL (Invalid >> argument) >> >> Looks like address is not 16K aligned. >> >> Ilya. >> >> 2012/11/1 Philippe Waroquiers <phi...@sk...**>: >> >>> On Wed, 2012-10-31 at 11:49 -0700, John Reiser wrote: >>> >>>> The ending address 0041de4fff is not the end of a 16KiB page: >>>> --3518:2:aspacem 9: RSVN 0041de4000-0041de4fff 4096 ----- >>>> SmFixed >>>> --3518:2:aspacem 10: 0041de5000-007fba3fff 989m >>>> Neither is the _next_ address 0041de5000 the beginning of a 16KiB page. >>>> >>> Interesting observation. >>> >>> It looks like the /proc/self/maps is already not aligned on 16Kb >>> pages. Is that normal if the kernel is configured to be 16Kb pages ? >>> What does >>> getconf -a | grep PAGESIZE >>> give ? >>> on gcc49, it gives >>> PAGESIZE 16384 >>> and there, the /proc/self/maps reading is properly page aligned. >>> >>> Maybe the kernel is not configured for 16Kb pages ? >>> >>> Otherwise, strace will help to see what is going on for failing >>> mmaps. >>> >>> Philippe >>> >>> >>> >>> ------------------------------**------------------------------** >>> ------------------ >>> Everyone hates slow websites. So do we. >>> Make your web apps faster with AppDynamics >>> Download AppDynamics Lite for free today: >>> http://p.sf.net/sfu/appdyn_**sfd2d_oct<http://p.sf.net/sfu/appdyn_sfd2d_oct> >>> ______________________________**_________________ >>> Valgrind-developers mailing list >>> Valgrind-developers@lists.**sourceforge.net<Val...@li...> >>> https://lists.sourceforge.net/**lists/listinfo/valgrind-**developers<https://lists.sourceforge.net/lists/listinfo/valgrind-developers> >>> >> ------------------------------**------------------------------** >> ------------------ >> Everyone hates slow websites. So do we. >> Make your web apps faster with AppDynamics >> Download AppDynamics Lite for free today: >> http://p.sf.net/sfu/appdyn_**sfd2d_oct<http://p.sf.net/sfu/appdyn_sfd2d_oct> >> ______________________________**_________________ >> Valgrind-developers mailing list >> Valgrind-developers@lists.**sourceforge.net<Val...@li...> >> https://lists.sourceforge.net/**lists/listinfo/valgrind-**developers<https://lists.sourceforge.net/lists/listinfo/valgrind-developers> >> >> > |
|
From: Philippe W. <phi...@sk...> - 2012-11-06 21:32:17
|
On Tue, 2012-11-06 at 17:25 +0700, Ilya Smelykh wrote: > /* configured default page size 4k */ > #define MIPS_PAGE_SHIFT 14 > > > /* configured memory alignment 2*PAGE_SIZE */ > #define SHM_ALIGNMENT 2*(1UL << MIPS_PAGE_SHIFT) What does getconf -a | grep PAGESIZE give ? What does strace -f valgrind .... gives ? (in particular about failing mmap) ? Philippe |
|
From: Ilya S. <ily...@gm...> - 2013-01-18 10:29:06
Attachments:
v.out
|
bin # /disk/strace -f -s 50 -o v.out valgrind -v -v -v -d -d -d --trace-syscalls=yes /disk/strace --1470:1:debuglog DebugLog system started by Stage 1, level 3 logging requested --1470:1:launcher no tool requested, defaulting to 'memcheck' --1470:2:launcher selecting platform for '/disk/strace' --1470:2:launcher selecting platform for '/disk/strace' --1470:2:launcher opened '/disk/strace' --1470:2:launcher read 4096 bytes from '/disk/strace' --1470:2:launcher selected platform 'mips32-linux' --1470:1:launcher selected platform 'mips32-linux' --1470:1:launcher launching /home/ilya/valgrind_mips/lib/valgrind/memcheck-mips32-linux --1470:1:debuglog DebugLog system started by Stage 2 (main), level 3 logging requested --1470:1:main Welcome to Valgrind version 3.9.0.SVN debug logging --1470:1:main Checking current stack is plausible --1470:1:main Checking initial stack was noted --1470:1:main Starting the address space manager --1470:2:aspacem sp_at_startup = 0x007fbef5f0 (supplied) --1470:2:aspacem minAddr = 0x0004000000 (computed) --1470:2:aspacem maxAddr = 0x007fbebfff (computed) --1470:2:aspacem cStart = 0x0004000000 (computed) --1470:2:aspacem vStart = 0x0041df8000 (computed) --1470:2:aspacem suggested_clstack_top = 0x007ebeffff (computed) --1470:2:aspacem <<< SHOW_SEGMENTS: Initial layout (5 segments, 0 segnames) --1470:2:aspacem 0: RSVN 0000000000-0003ffffff 64m ----- SmFixed --1470:2:aspacem 1: 0004000000-0041df7fff 989m --1470:2:aspacem 2: RSVN 0041df8000-0041dfbfff 16384 ----- SmFixed --1470:2:aspacem 3: 0041dfc000-007fbebfff 989m --1470:2:aspacem 4: RSVN 007fbec000-00ffffffff 2052m ----- SmFixed --1470:2:aspacem >>> --1470:2:aspacem Reading /proc/self/maps --1470:2:aspacem <<< SHOW_SEGMENTS: With contents of /proc/self/maps (15 segments, 1 segnames) --1470:2:aspacem ( 0) /disk/lib/valgrind/memcheck-mips32-linux --1470:2:aspacem 0: RSVN 0000000000-00003fffff 4194304 ----- SmFixed --1470:2:aspacem 1: FILE 0000400000-0000403fff 16384 r---- d=0x803 i=18989649 o=0 (0) --1470:2:aspacem 2: RSVN 0000404000-0003ffffff 59m ----- SmFixed --1470:2:aspacem 3: 0004000000-0037ffffff 832m --1470:2:aspacem 4: FILE 0038000000-003839ffff 3801088 r-x-- d=0x803 i=18989649 o=65536 (0) --1470:2:aspacem 5: 00383a0000-00383abfff 49152 --1470:2:aspacem 6: FILE 00383ac000-00383bbfff 65536 rw--- d=0x803 i=18989649 o=3850240 (0) --1470:2:aspacem 7: ANON 00383bc000-0038e83fff 10m rwx-- --1470:2:aspacem 8: 0038e84000-0041df7fff 143m --1470:2:aspacem 9: RSVN 0041df8000-0041dfbfff 16384 ----- SmFixed --1470:2:aspacem 10: 0041dfc000-007fbcbfff 989m --1470:2:aspacem 11: ANON 007fbcc000-007fbeffff 147456 rwx-- --1470:2:aspacem 12: RSVN 007fbf0000-007fff3fff 4210688 ----- SmFixed --1470:2:aspacem 13: ANON 007fff4000-007fff7fff 16384 r-x-- --1470:2:aspacem 14: RSVN 007fff8000-00ffffffff 2048m ----- SmFixed --1470:2:aspacem >>> --1470:1:main Address space manager is running --1470:1:main Starting the dynamic memory manager --1470:1:mallocfr newSuperblock at 0x41DFC000 (pszB 4194288) owner VALGRIND/tool --1470:1:mallocfr deferred_reclaimSuperblock at 0x41DFC000 (pszB 4194288) (prev 0x0) owner VALGRIND/tool --1470:1:main Dynamic memory manager is running --1470:1:main Initialise m_debuginfo --1470:1:main VG_(libdir) = /home/ilya/valgrind_mips/lib/valgrind --1470:1:main Getting launcher's name ... --1470:1:main ... /bin/valgrind --1470:1:main Get hardware capabilities ... --1470:1:cache Could not autodetect cache info --1470:1:main ... arch = MIPS32, hwcaps = MIPS-baseline --1470:1:main Getting the working directory at startup --1470:1:main ... /disk/valgrind_mips/bin --1470:1:main Split up command line --1470:1:main (early_) Process Valgrind's command line options --1470:1:main Create initial image --1470:1:initimg Loading client valgrind: mmap(0x400000, 294912) failed in UME with error 22 (Invalid argument). valgrind: this can be caused by executables with very large text, data or bss segments. |
|
From: Dejan J. <dej...@rt...> - 2013-01-18 13:41:23
Attachments:
mipsisa32r2el_gcc.diff
|
On 01/18/2013 11:28 AM, Ilya Smelykh wrote: > Please find attached logs. > > > 2012/11/7 Philippe Waroquiers <phi...@sk... > <mailto:phi...@sk...>> > > On Tue, 2012-11-06 at 17:25 +0700, Ilya Smelykh wrote: > > /* configured default page size 4k */ > > #define MIPS_PAGE_SHIFT 14 > > > > > > /* configured memory alignment 2*PAGE_SIZE */ > > #define SHM_ALIGNMENT 2*(1UL << MIPS_PAGE_SHIFT) > What does > getconf -a | grep PAGESIZE > give ? > > What does > strace -f valgrind .... > gives ? (in particular about failing mmap) ? > > Philippe > > > > Hi Ilya, Cal you apply this patch, reconfigure and recompile V again. Regards, Dejan |
|
From: Ilya S. <ily...@gm...> - 2013-01-18 14:06:12
|
Hi Dejan, It seems this patch solved my problem! Thank you very much, I will continue to testing V on my target system, and will report if any problems will be found. Thanks a lot. Ilya. 2013/1/18 Dejan Jevtic <dej...@rt...> > On 01/18/2013 11:28 AM, Ilya Smelykh wrote: > > Please find attached logs. > > > 2012/11/7 Philippe Waroquiers <phi...@sk...> > >> On Tue, 2012-11-06 at 17:25 +0700, Ilya Smelykh wrote: >> > /* configured default page size 4k */ >> > #define MIPS_PAGE_SHIFT 14 >> > >> > >> > /* configured memory alignment 2*PAGE_SIZE */ >> > #define SHM_ALIGNMENT 2*(1UL << MIPS_PAGE_SHIFT) >> What does >> getconf -a | grep PAGESIZE >> give ? >> >> What does >> strace -f valgrind .... >> gives ? (in particular about failing mmap) ? >> >> Philippe >> >> >> >> > Hi Ilya, > > Cal you apply this patch, reconfigure and recompile V again. > > Regards, > Dejan > |
|
From: Petar J. <mip...@gm...> - 2013-01-21 01:50:15
|
On Fri, Jan 18, 2013 at 3:05 PM, Ilya Smelykh <ily...@gm...> wrote: > Hi Dejan, > It seems this patch solved my problem! > Thank you very much, I will continue to testing V on my target system, and > will report if any problems will be found. > > Thanks a lot. > Ilya. Can you try to download the trunk version now (or any r13249+) and see if it works for you? Petar |