Kevin, As per your suggestion, I checked the __log_buf memory using CCS , it had the below contents
<5>Linux version 2.6.28-rc8-davinci1 (gcc version 3.4.3 (MontaVista 3.4.3-25.0.104.0600975 2006-07-06)) #4 PREEMPT Tue Dec 16 13:35:43 EST 2008 <4>CPU: ARM926EJ-S [41069265] revision 5 (ARMv5TEJ), cr=00053177 <4>CPU: VIVT data cache, VIVT instruction cache <4>Machine: DaVinci DM646x EVM <4>Memory policy: ECC disabled, Data cache writeback <7>On node 0 totalpages: 32768 <7>free_area_init_node: node 0, pgdat c02ebf34, node_mem_map c031a000 <7> DMA zone: 256 pages used for memmap <7> DMA zone: 0 pages reserved <7> DMA zone: 32512 pages, LIFO batch:7 <7> Normal zone: 0 pages used for memmap <7> Movable zone: 0 pages used for memmap <6>DaVinci DM6467 variant 0x0 After this there are no messages in the log. On compiling with CONFIG_DEBUG_LL, there is no serial output and CCS is not able to halt the CPU for this case, so the kernel printk messages could not be observed. I will check the same for the kernels before and after the commit(6c76c1971a42eb96448c10f914e444e35049cdcd) you have mentioned and will post the output. Thanks, Chaithrika ________________________________________ From: Kevin Hilman [[email protected]] Sent: Monday, December 15, 2008 11:31 PM To: Subrahmanya, Chaithrika Cc: [email protected] Subject: Re: linux-2.6.28-rc8 broken for DM646x EVM "Subrahmanya, Chaithrika" <[email protected]> writes: > The latest kernel seems to be broken for DM646x EVM. > It does not boot on the DM646x EVM but runs on DM644x EVM. It has not been > tested on DM355. > The kernel image was built using both the generic and DM646x specific > defconfig files. > > Compiling with low level kernel debug messages did not help either as there > was no serial output after the > "Uncompressing > Linux............................................................................................... > done, booting the kernel." message. > Chaithrika, I noticed this too, though with CONFIG_DEBUG_LL, I see garbage on my serial console. If you checkout at this commit: 6c76c1971a42eb96448c10f914e444e35049cdcd, dm6467 boots again. Checking out the next commit, which is a merge with upstream, it no longer works for me. A quick look a the diff between the two commits, I didn't see anything obvious changing in DaVinci specific code. I don't have a working CCS for this platform, but if you have a working CCS, it might be worth debugging where the code actually is. On quick way is to just dump the kernel printk log. Look for the address of the symbol __log_buf in System.map. Then using CCS, halt the kernel and dump memory at that location in ASCII. You should then see the kernel printk messages. Kevin_______________________________________________ Davinci-linux-open-source mailing list [email protected] http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
