June 2011 - kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

Changes in block layer for IOScheduler Insertion selection
by mani Feb. 6, 2012

Feb. 6, 2012

Dear All, I am working on linux kernel 2.6.32.9 tegra NVIDIA board. I am getting ~8MBps speed of the Nand disk if i use hdparm hdparm -t /dev/mtdblock3 i made changes in block layer of kernel as below:- block/blk-core.c static inline void add_request(struct request_queue *q, struct request *req) { drive_stat_acct(req, 1); /* * elevator indicated where it wants this request to be * inserted at elevator_merge time */ __elv_add_request(q, req, ELEVATOR_INSERT_BACK, 0); //__elv_add_request(q, req, ELEVATOR_INSERT_SORT, 0); } changed ELEVATOR_INSERT_SORT to ELEVATOR_INSERT_BACK it improves my NAND speed to 10MBps. I am using "noop" I/O scheduler. Will this change have any adverse effect in kernel ? or any other side effect as far as i am using only Nand no Hard disk. Thanks Mani

3 4

Can i allocate 4GB virtual addresses (more than a certain limit) using vmalloc?
by sandeep kumar Aug. 22, 2011

Aug. 22, 2011

Hi all, The following link gives the memory map for the arm architecture. http://www.arm.linux.org.uk/developer/memory.txt I have the following doubts.. 1) Any chipset(based on arm) manufacturer(qualcom,samsung..) should follow the same memory map. Is it hardly constrained or can be changed? Where are this constraints are implemented in the kernel source tree? 2) while i was student, i read in OS concepts that, "Virtual memory gives an illusion to a process, that it has always a larger continuous address space (even more than RAM) available to it." So i thought i could allocate howmuch ever memory i want. But seeing the above link,i observed there is some limitation in the address space created by the vmalloc(). So i m now thinking that vmalloc has some limit. Please make me clear these things.... With regards, Sandeep Kumar Anantapalli,

6 7

SRIO maintenance read hangs (mpc8548)
by shrikant hegde Aug. 4, 2011

Aug. 4, 2011

Hi all, I am working on srio; When i tried to send maintenance packet to switch it hangs. Symptom when i tried to read DIDCAR of switch,( i.e read 0x8000_0000 ), or HDIDLOCK (0x8000_0068 ), execution of code hangs; if i press break in code warrior its coming out with an error. register value dump EPWISR 0x80000000 -- physical or logical/transport error interrupt LTLEDCSR 0x08000000 -- illegal transaction decode, received request/response packet LTLCCCSR 0x82000000 Below are my system detail: Processor mpc8548 and Tsi578 switch. Processor is connected to switch on same board.I have written standalone code (no OS) to exercise srio, using code warrior. Processor side configurations Its configured to 1x, 2.5gbs, host. law and tlb are configured. 0x8000_0000 0x8FFF_FFFF maintenance window configured 0x8000_0000 0x8000_0FFF (4k) switch side configuration 1x,2.5gbs, swap disabled (configured during POR) Processor side link training bit is setting. Am I missing any configuration , in switch side or processor side ? Any pointer is very helpful. Forgive me if its a wrong forum to post srio queries. thanks shrikant -- View this message in context: http://old.nabble.com/SRIO-maintenance-read-hangs-%28mpc8548%29-tp30949266p… Sent from the kernelnewbies mailing list archive at Nabble.com.

2 1

block size check in __getblk_slow
by piyush moghe July 1, 2011

July 1, 2011

Hi All, I am new to linux kernel code and interested in filesystems and block I/O. While going through the block read code ( in __bread function ) I found that when we try to read the block ( __getblk ) we first check for the block buffer page in page cache ( using __find_get_block function ) and if not we do a slow path ( using __getblk_slow function which checks page in cache again and if not found again then allocate a buffer page in cache ). Now my problem is, as part of __getblk_slow we compare block size with bdev logical block size and also check that block size should not be less than 512 and not greater than PAGE_SIZE ( 4K, for PAGE_SHIFT set to 13 ) although these all checks can be moved to upper level function i.e __getblk and that too before calling __find_get_block ( to find block in page cache ) so as to avoid looking for block in page cache if the size is invalid. Please let me know your comments on this. Regards, Piyush Moghe

2 2

checking atags for boot parameters
by Christopher Harvey June 30, 2011

June 30, 2011

I'm suspicious that the bootloader I'm using isn't passing atags to the kernel properly. I'm getting a prefetch exception right after the "uncompressing linux....done" message, does it make sense that this could happen because mem=128M isn't getting passed properly? Basically what I want to do is have the kernel print out the boot parameters along with the "uncompressing linux" message. thanks, Christopher

3 3

How to find Bluetooth device type
by पारस June 30, 2011

June 30, 2011

Hi, I recently started working on Bluetooth. I write a c program to search nearby bluetooth devices. In search output, I get different types of bluetooth devices e.g. mobile phones, laptops. I want to display only mobile phones's bdaddress. Where do I get this information form blueZ stack. In which structure the device type is store. I am working on Ubuntu-10.4 and using blueZ-4.94. Thank you in advance.

4 5

Debugging a custom kernel
by Apelete Seketeli June 30, 2011

June 30, 2011

Hello, I am working on a custom kernel, and I would like to add the necessary support to enable it to boot with qemu. In order to achieve that I am trying to debug it inside qemu by attaching a gdb to it. I still can't figure where the boot process is getting stuck with step-by-step execution, but it seems that the last function called is "delay_loop" from arch/x86/lib/delay.c. That function contains some assembly code, does someone know what it is supposed to do ? Beside, do you have any advice on the way I should proceed to get the necessary information to port the kernel ? Thanks. -- Apelete

3 4

Re: hlist_add_head
by piyush moghe June 30, 2011

June 30, 2011

On Thu, Jun 30, 2011 at 2:47 PM, piyush moghe <pmkernel(a)gmail.com> wrote: > AFAIK hlist_add_head is a function to add a node at first position in > collision chain at particular head ( hash bucket ) > > For ex. whenever a new inode is allocated the hash value of that inode is > calculated which acts as an index in inode_hashtable > which subsequently provides and hash bucket ( head ) for that inode. Now > this inode attached at the first position in collision chain for the > calculated hash bucket pushing the current first inode to second position. > > For further clarification have a look at inode_add_to_lists() and > __inode_add_to_lists() functions. > > Regards, > Piyush > > > On Tue, Jun 28, 2011 at 8:05 PM, luca ellero <lroluk(a)gmail.com> wrote: > >> Hi everybody, >> can someone please explain what is the role of hlist_add_head (used in >> in hash lists). Or maybe provide some links with good examples of its >> use. I had a look at fs/dcache.c but frankly I can't understand it very >> well. >> Thanks in advance >> Luca >> >> _______________________________________________ >> Kernelnewbies mailing list >> Kernelnewbies(a)kernelnewbies.org >> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies >> > >

1 0

finding kernel jump address after "decompressing linux"
by Christopher Harvey June 30, 2011

June 30, 2011

I'm trying to figure out what physical address the kernel jumps to after "Uncompressing Linux... done, booting the kernel.". IIRC, there are two parts to a kernel image, one compressed part and one uncompressed. The uncompressed code decompresses the compressed part and puts it into memory then jumps to it. I'm using an ARM kernel, version 2.6.38. thanks, Chris

3 2

Kernel log mechanism: How to log messages in a different file
by amit mehta June 30, 2011

June 30, 2011

Hi, I've written some kernel modules using kprobes infrastructure to grab some more information on network packets. I've lots of printk which by default show up in syslog file(/var/log/messages), I was wondering if there are some method to redirect my printkS to appear in somewhere else(say /var/log/netlog). On a busy system, it gets little cluttered with other logs. Of-course i can grep for my own custom messages but it would be lot more convenient if i can redirect my messages in a separate log file. Considering above can be done, then what about dmesg ? will they still show up in dmesg output ? Regards, Amit

5 5