doubt Shared memory problem
hi folks here i am fire fighting with a shared memory error in linux platforms i developed an application in a 32 bit machine, and i ported it in to 64 bit machine it was working porperly in 32 bit machine. unfortunately it is not working properly in 64 bit maching giving an error called no space left on the device the application will take html files as its input (it may be 10k, 30k and so on) but in 64 bit maching its only taking 10k file thats all if it goes more than that i will crash i dont know what happened please get back to me as early as possible regards arun
Re: doubt Shared memory problem
On Mon, 6 Jun 2005, ARUN wrote: hi folks here i am fire fighting with a shared memory error in linux platforms i developed an application in a 32 bit machine, and i ported it in to 64 bit machine it was working porperly in 32 bit machine. unfortunately it is not working properly in 64 bit maching Hi Arun, I assume you are talking about C or C++ here. Am I correct? You may have many problems there, and not just one. A good place to start is to compile your code with strict warnings, and make sure you do not have any (-Werror). For example, using gcc, use -Wall -Wconversion -Wcast-qual -Wpointer-arith -Wsign-compare -Wcast-align and some other stuff may help as well. If not, you will have to be more specific. Orna. -- Orna Agmon http://ladypine.org/ http://haifux.org/~ladypine/ ICQ: 348759096 = To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: doubt Shared memory problem
More details are needed: 1. In which programming language and for which environment was your application developed? 2. If it was developed in C/C++, Java or something similar, what data types are being used by it and what are their sizes? 3. Are there any built-in arbitrary limits on sizes of data structures? 4. Do some of those limits depend upon the size of integers? --- Omer On Mon, 2005-06-06 at 13:43 +0530, ARUN wrote: hi folks here i am fire fighting with a shared memory error in linux platforms i developed an application in a 32 bit machine, and i ported it in to 64 bit machine it was working porperly in 32 bit machine. unfortunately it is not working properly in 64 bit maching giving an error called no space left on the device the application will take html files as its input (it may be 10k, 30k and so on) but in 64 bit maching its only taking 10k file thats all if it goes more than that i will crash i dont know what happened please get back to me as early as possible regards arun -- One does not make peace with enemies. One makes peace with former enemies. My own blog is at http://www.livejournal.com/users/tddpirate/ My opinions, as expressed in this E-mail message, are mine alone. They do not represent the official policy of any organization with which I may be affiliated in any way. WARNING TO SPAMMERS: at http://www.zak.co.il/spamwarning.html = To unsubscribe, send mail to [EMAIL PROTECTED] with the word unsubscribe in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
A memory problem ?
I have been using Linux on my Laptop since a couple of months. I first installed the kernel 2.2.12 (Red Hat 6.1) which I recently updated to 2.2.16-3. I decided to update the kernel since I thought that Linux was using too much memory and that there might be a problem (I read once that the kernel 2.2.12 was known to have some memory leach). Unfortunatly, updating my kernel didn't change anything. My question is the following : does my Linux indeed use too much memory (and in that case have you got any idea why) or is every thing normal ? Here is how much memories Linux takes (by the way, my kernel of normal size : its bzImage is around 620 K) I boot and I log in: total used free sharedbuffers cached Mem:127948 28820 99128 11476 3212 16912 -/+ buffers/cache: 8696 119252 Swap:68000 0 68000 Then I start X: total used free sharedbuffers cached Mem:127928 41536 86392 17920 3848 24868 -/+ buffers/cache: 12820 115108 Swap:68000 0 68000 Then I launch Netscape (4.73): total used free sharedbuffers cached Mem:127928 64464 63464 32920 4284 39580 -/+ buffers/cache: 20600 107328 Swap:68000 0 68000 Then I start Emacs: total used free sharedbuffers cached Mem:127928 69868 58060 35864 4440 42640 -/+ buffers/cache: 22788 105140 Swap:68000 0 68000 Thanks a lot for your help ! Emmanuel. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Kernel/Memory problem.
I had the same crash on nonBrandName stations used whith Squid, but never on
Dell or
Compaq. It seems like memory physical defect.
Mike Almogy wrote:
Hi list.
I just had my P233 512 MB RAM 2 9.1GB 10K RPM SCSI disk box crushed !.
In the message log i got the following lines :
Jun 24 03:03:00 proxyint kernel: Uhhuh. NMI received. Dazed and confused,
but trying to continue
Jun 24 03:03:00 proxyint kernel: You probably have a hardware problem with
your RAM chips
Jun 24 03:03:00 proxyint kernel: Unable to handle kernel paging request at
virtual address 99ce6fc8
Jun 24 03:03:00 proxyint kernel: current-tss.cr3 = 00101000, %cr3 =
00101000
Jun 24 03:03:00 proxyint kernel: *pde =
Jun 24 03:03:00 proxyint kernel: Oops:
Jun 24 03:03:00 proxyint kernel: CPU:0
Jun 24 03:03:00 proxyint kernel: EIP:0010:[kmem_cache_free+61/353]
Jun 24 03:03:00 proxyint kernel: EFLAGS: 00010082
Jun 24 03:03:00 proxyint kernel: eax: 00bc ebx: d9ce69c0 ecx:
99ce6fc0 edx: d9ce6a7c
Jun 24 03:03:00 proxyint kernel: esi: dff9b980 edi: 0286 ebp:
esp: d9c79e38
Jun 24 03:03:00 proxyint kernel: ds: 0018 es: 0018 ss: 0018
Jun 24 03:03:00 proxyint kernel: Process squid (pid: 17995, process nr: 32,
stackpage=d9c79000)
Jun 24 03:03:00 proxyint kernel: Stack: d9ce6a1c d9ce6a7c d16e562c
c014c21d dff9b980 d9ce69c0 d9ce69c0
Jun 24 03:03:00 proxyint kernel:c014c2c9 d9ce69c0 df478444 c8ad80c0
c0160e08 d9ce69c0 df478400 d4e957fc
Jun 24 03:03:00 proxyint kernel:c8ad80c0 0003 8218 d4e957fc
df478400 d4e957fc c016e0c1 df478400
Jun 24 03:03:00 proxyint kernel: Call Trace: [kfree_skbmem+50/61]
[__kfree_skb+161/167] [tcp_close+208/601] [inet_release+122/130]
[sock_release+31/80] [sock_close+50/57] [__fput+31/69]
Jun 24 03:03:00 proxyint kernel:[fput+23/68] [filp_close+80/89]
[do_exit+288/616] [do_signal+487/601] [force_sig_info+121/129]
[force_sig+17/21] [do_page_fault+807/883] [error_code+45/52]
Jun 24 03:03:00 proxyint kernel:[signal_return+20/24]
Jun 24 03:03:00 proxyint kernel: Code: 8b 69 08 81 fd 2b 2f c3 a5 0f 85 d0
00 00 00 8b 69 0c 85 ed
Jun 24 03:03:00 proxyint kernel: Unable to handle kernel paging request at
virtual address 5d9c
Jun 24 03:03:00 proxyint kernel: current-tss.cr3 = 0ca43000, %cr3 =
0ca43000
Jun 24 03:03:00 proxyint kernel: *pde =
Jun 24 03:03:00 proxyint kernel: Oops:
Jun 24 03:03:00 proxyint kernel: CPU:0
Jun 24 03:03:00 proxyint kernel: EIP:0010:[fput+5/68]
Jun 24 03:03:00 proxyint kernel: EFLAGS: 00010282
Jun 24 03:03:00 proxyint kernel: eax: ebx: 5d80 ecx:
edx:
Jun 24 03:03:00 proxyint kernel: esi: edi: ebp:
1000 esp: dab3dfac
Jun 24 03:03:00 proxyint kernel: ds: 0018 es: 0018 ss: 0018
Jun 24 03:03:00 proxyint kernel: Process dnsserver (pid: 18003, process nr:
36, stackpage=dab3d000)
Jun 24 03:03:00 proxyint kernel: Stack: 5d80 dab3c000 401498c0 bb58
ba64 c010a0d4 40015000
Jun 24 03:03:00 proxyint kernel:1000 401498c0 bb58 ba64
0003 002b 002b 0003
Jun 24 03:03:00 proxyint kernel:40101ad4 0023 0202 ba4c
002b
Jun 24 03:03:00 proxyint kernel: Call Trace: [system_call+52/56]
Jun 24 03:03:00 proxyint kernel: Code: 8b 43 1c 48 75 34 53 e8 8b 96 00 00
53 e8 8b ef ff ff c7 43
Jun 24 03:03:00 proxyint kernel: swap_duplicate: entry 4007, offset
exceeds max
Jun 24 03:03:37 proxyint squid[458]: Squid Parent: child process 18062
started
Jun 24 03:03:00 proxyint kernel: VM: killing process dnsserver
Jun 24 03:03:37 proxyint kernel: swap_free: offset exceeds max
Jun 24 03:03:37 proxyint kernel: swap_free: swap-space map bad (entry
0007)
Jun 24 03:03:37 proxyint kernel: swap_free: offset exceeds max
Jun 24 03:03:37 proxyint kernel: free_one_pmd: bad directory entry 4014
Jun 24 03:03:37 proxyint kernel: free_one_pmd: bad directory entry bfff
Jun 24 03:03:37 proxyint kernel: free_one_pmd: bad directory entry 100f
Jun 24 03:03:37 proxyint kernel: magic (corrupt) (name=size-128)
Jun 24 03:03:37 proxyint kernel: kmem_alloc: Bad slab magic (corrupt)
(name=size-128)
Jun 24 03:05:08 proxyint last message repeated 292 times
Jun 24 03:05:37 proxyint kernel: magic (corrupt) (name=size-128)
Jun 24 03:05:37 proxyint kernel: kmem_alloc: Bad slab magic (corrupt)
(name=size-128)
Jun 24 03:06:37 proxyint last message repeated 71 times
Jun 24 03:06:37 proxyint kernel: kmem_alloc: Bad slab magic (magic (corrupt)
(name=size-128)
Jun 24 03:06:37 proxyint kernel: kmem_alloc: Bad slab magic (corrupt)
(name=size-128)
Jun 24 03:06:50 proxyint last message repeated 292 times
Jun 24 03:12:37 proxyint kernel: d slab magic (corrupt) (name=size-128)
Jun 24 03:14:37 proxyint kernel: kmem_alloc: Bad slab magic (corrupt)
(name=size-128)
I notice that there is a message there about maybe a
Kernel/Memory problem.
Hi list. I just had my P233 512 MB RAM 2 9.1GB 10K RPM SCSI disk box crushed !. In the message log i got the following lines : Jun 24 03:03:00 proxyint kernel: Uhhuh. NMI received. Dazed and confused, but trying to continue Jun 24 03:03:00 proxyint kernel: You probably have a hardware problem with your RAM chips Jun 24 03:03:00 proxyint kernel: Unable to handle kernel paging request at virtual address 99ce6fc8 Jun 24 03:03:00 proxyint kernel: current-tss.cr3 = 00101000, %cr3 = 00101000 Jun 24 03:03:00 proxyint kernel: *pde = Jun 24 03:03:00 proxyint kernel: Oops: Jun 24 03:03:00 proxyint kernel: CPU:0 Jun 24 03:03:00 proxyint kernel: EIP:0010:[kmem_cache_free+61/353] Jun 24 03:03:00 proxyint kernel: EFLAGS: 00010082 Jun 24 03:03:00 proxyint kernel: eax: 00bc ebx: d9ce69c0 ecx: 99ce6fc0 edx: d9ce6a7c Jun 24 03:03:00 proxyint kernel: esi: dff9b980 edi: 0286 ebp: esp: d9c79e38 Jun 24 03:03:00 proxyint kernel: ds: 0018 es: 0018 ss: 0018 Jun 24 03:03:00 proxyint kernel: Process squid (pid: 17995, process nr: 32, stackpage=d9c79000) Jun 24 03:03:00 proxyint kernel: Stack: d9ce6a1c d9ce6a7c d16e562c c014c21d dff9b980 d9ce69c0 d9ce69c0 Jun 24 03:03:00 proxyint kernel:c014c2c9 d9ce69c0 df478444 c8ad80c0 c0160e08 d9ce69c0 df478400 d4e957fc Jun 24 03:03:00 proxyint kernel:c8ad80c0 0003 8218 d4e957fc df478400 d4e957fc c016e0c1 df478400 Jun 24 03:03:00 proxyint kernel: Call Trace: [kfree_skbmem+50/61] [__kfree_skb+161/167] [tcp_close+208/601] [inet_release+122/130] [sock_release+31/80] [sock_close+50/57] [__fput+31/69] Jun 24 03:03:00 proxyint kernel:[fput+23/68] [filp_close+80/89] [do_exit+288/616] [do_signal+487/601] [force_sig_info+121/129] [force_sig+17/21] [do_page_fault+807/883] [error_code+45/52] Jun 24 03:03:00 proxyint kernel:[signal_return+20/24] Jun 24 03:03:00 proxyint kernel: Code: 8b 69 08 81 fd 2b 2f c3 a5 0f 85 d0 00 00 00 8b 69 0c 85 ed Jun 24 03:03:00 proxyint kernel: Unable to handle kernel paging request at virtual address 5d9c Jun 24 03:03:00 proxyint kernel: current-tss.cr3 = 0ca43000, %cr3 = 0ca43000 Jun 24 03:03:00 proxyint kernel: *pde = Jun 24 03:03:00 proxyint kernel: Oops: Jun 24 03:03:00 proxyint kernel: CPU:0 Jun 24 03:03:00 proxyint kernel: EIP:0010:[fput+5/68] Jun 24 03:03:00 proxyint kernel: EFLAGS: 00010282 Jun 24 03:03:00 proxyint kernel: eax: ebx: 5d80 ecx: edx: Jun 24 03:03:00 proxyint kernel: esi: edi: ebp: 1000 esp: dab3dfac Jun 24 03:03:00 proxyint kernel: ds: 0018 es: 0018 ss: 0018 Jun 24 03:03:00 proxyint kernel: Process dnsserver (pid: 18003, process nr: 36, stackpage=dab3d000) Jun 24 03:03:00 proxyint kernel: Stack: 5d80 dab3c000 401498c0 bb58 ba64 c010a0d4 40015000 Jun 24 03:03:00 proxyint kernel:1000 401498c0 bb58 ba64 0003 002b 002b 0003 Jun 24 03:03:00 proxyint kernel:40101ad4 0023 0202 ba4c 002b Jun 24 03:03:00 proxyint kernel: Call Trace: [system_call+52/56] Jun 24 03:03:00 proxyint kernel: Code: 8b 43 1c 48 75 34 53 e8 8b 96 00 00 53 e8 8b ef ff ff c7 43 Jun 24 03:03:00 proxyint kernel: swap_duplicate: entry 4007, offset exceeds max Jun 24 03:03:37 proxyint squid[458]: Squid Parent: child process 18062 started Jun 24 03:03:00 proxyint kernel: VM: killing process dnsserver Jun 24 03:03:37 proxyint kernel: swap_free: offset exceeds max Jun 24 03:03:37 proxyint kernel: swap_free: swap-space map bad (entry 0007) Jun 24 03:03:37 proxyint kernel: swap_free: offset exceeds max Jun 24 03:03:37 proxyint kernel: free_one_pmd: bad directory entry 4014 Jun 24 03:03:37 proxyint kernel: free_one_pmd: bad directory entry bfff Jun 24 03:03:37 proxyint kernel: free_one_pmd: bad directory entry 100f Jun 24 03:03:37 proxyint kernel: magic (corrupt) (name=size-128) Jun 24 03:03:37 proxyint kernel: kmem_alloc: Bad slab magic (corrupt) (name=size-128) Jun 24 03:05:08 proxyint last message repeated 292 times Jun 24 03:05:37 proxyint kernel: magic (corrupt) (name=size-128) Jun 24 03:05:37 proxyint kernel: kmem_alloc: Bad slab magic (corrupt) (name=size-128) Jun 24 03:06:37 proxyint last message repeated 71 times Jun 24 03:06:37 proxyint kernel: kmem_alloc: Bad slab magic (magic (corrupt) (name=size-128) Jun 24 03:06:37 proxyint kernel: kmem_alloc: Bad slab magic (corrupt) (name=size-128) Jun 24 03:06:50 proxyint last message repeated 292 times Jun 24 03:12:37 proxyint kernel: d slab magic (corrupt) (name=size-128) Jun 24 03:14:37 proxyint kernel: kmem_alloc: Bad slab magic (corrupt) (name=size-128) I notice that there is a message there about maybe a problem with my RAM. Is there any tests that i can run in order to verify that the RAM is really the thing that made the kernel to crush ? I also got some screen messages that said that the swap/paging mechanism stop working and that for some
Re: Kernel/Memory problem.
On Sat, 24 Jun 2000, Mike Almogy wrote: Here the thing, the server is Compaq 1200. It runs squid from the day we got it. But with lots of problems. well, Compaq should be able to help. they give warrenty and service on the hardware thy sell, and they openly support Linux. Do you think that i need a better hardware ? I don't know the Compaq in question. but it sounds like you should consider moving the cache repository over to 2-3 seperate drives (Squid does the balancing in a smarter way than any RAID can). Format them with ReiserFS, it will add stability and speed too. i'm thinking to get 6 disks box with duel CPU and 1 GB of ram. now you're talking :-) 6 are better than 3, all depends on your needs. BTW, i had 4 swap partitions due to the sad truth that Linux did not used the swap at all, i thought that it might be because i gave it much larger swap then it can handle. no. Linux once had the limitation of 128 meg swap partitions, this is no longer the case, I think it was raised to 2 gigs. what you are experiancing is really squid being efficiant. squid will cache as much as possible in RAM for fast access to pages, and what it can't, it will use the disk for. if it allowed itself to swap out parts of the RAM it uses to the disk, you would lose the advantage of speed in accessing RAM chips instead of disks, and it would not have been as well layed out as the info it now reserves there. therefore it was decided in the last major version of squid to stop using swap memory altogether, and just use live RAM. the machine should not need to swap out anything unless you are running anything other than squid. Plus if i need to reformat the swap space then i can do it by parts and not to leave the server with out swap at all. just keep it on a seperate drive, that's all. a big cache server can boot from a dumb IDE drive or regular SCSI and cache on Ultra SCSI drives. I would have set it up that way. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: memory problem
i have an old 133MHZ pentium with 32Mb of ram. my linux (red-hat 6.0 kernel 2.2.5-15) detects only 13-14MB. i solved this problem with linux mem=32M at lilo prompt. it's seems to work fine but when i try to run the command startx it gives me an error message: segmentation fault and halt's. is there any other way to solve this. It sounds as if you have memory hole enabled (usually done for isa vga cards). If you don't need the hole, disable it in the bios. If you do have an isa vga card you have a few options: 1. disable the memory hole, so the card can't be used in linear mode, so XF86 will only run at 8bpp. svgalib (and maybe commercial X servers) can still use high color modes. 2. Enable the memory hole and use only 14MB of ram. 3. Enable the hole, and either upgrade to 2.3, or look for a e820 patch for 2.2 kernel. -- Matan Ziv-Av. [EMAIL PROTECTED] = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: memory problem
Some BIOSes used to have an option for 15-16 memory hole. Since your kernel initially report on no more then 15M ram I would check if you have such an option and if it is not enabled. Hi all i have an old 133MHZ pentium with 32Mb of ram. my linux (red-hat 6.0 kernel 2.2.5-15) detects only 13-14MB. i solved this problem with linux mem=32M at lilo prompt. it's seems to work fine but when i try to run the command startx it gives me an error message: segmentation fault and halt's. is thereany other way to solve this. please help.
Re: memory problem
- Original Message - From: Matan Ziv-Av [EMAIL PROTECTED] To: úåîø àééæðîï [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Sent: Sunday, January 09, 2000 11:15 AM Subject: Re: memory problem i have an old 133MHZ pentium with 32Mb of ram. my linux (red-hat 6.0 kernel 2.2.5-15) detects only 13-14MB. i solved this problem with linux mem=32M at lilo prompt. it's seems to work fine but when i try to run the command startx it gives me an error message: segmentation fault and halt's. is there any other way to solve this. It sounds as if you have memory hole enabled (usually done for isa vga cards). If you don't need the hole, disable it in the bios. If you do have an isa vga card you have a few options: 1. disable the memory hole, so the card can't be used in linear mode, so XF86 will only run at 8bpp. svgalib (and maybe commercial X servers) can still use high color modes. 2. Enable the memory hole and use only 14MB of ram. 3. Enable the hole, and either upgrade to 2.3, or look for a e820 patch for 2.2 kernel. -- Matan Ziv-Av. [EMAIL PROTECTED] Yep, that was the problem. i disabled it and it works fine now. thank you all. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
