User talk:Masterzorag
As far as I know, I'm the only coding OpenCL on the Cell here, if someone want to test something be warned that due some spufs changes that ppc-kernel-devs are (maybe) trying to fix, latest 3.3/3.4/3.5 branches falls into 'possible circular locking dependency detected' and slowdown runtime.
- It's stable until 3.2 branch.
- Even disabling lock debugging it slowdowns without warnings, it happens even with OpenCL samples from IBM.
http://permalink.gmane.org/gmane.linux.ports.ppc.embedded/50547
Latest tested kernels:
- 3.2.55 works fine
# ./perlin OpenCL took 22.496168 seconds to compute 1000 frames. Pixel Rate = 46.611316 Mpixels/sec, Frame Rate = 44.452015 frames/sec Host code took 12.620616 seconds to compute 10 frames. Pixel Rate = 0.830844 Mpixels/sec, Frame Rate = 0.792354 frames/sec OpenCL provided a 56.101182 speedup
- 3.3.3/3.4.6/3.5.3 falls into 'possible circular locking dependency detected' and slowdown runtime
Here the slowdown effect:
# ./perlin OpenCL took 93.280273 seconds to compute 1000 frames. Pixel Rate = 11.241133 Mpixels/sec, Frame Rate = 10.720380 frames/sec Host code took 12.948244 seconds to compute 10 frames. Pixel Rate = 0.809821 Mpixels/sec, Frame Rate = 0.772305 frames/sec OpenCL provided a 13.881010 speedup
In this specific case time spent is 4x to do the same thing!
When program runs something is going weird, e.g. in my program I'm used to query an OpenCL builtin function to tell me how many available SPEs there are, and its reply 8.
Using spu_base.enum_shared=1 parameter it should reply 7, so seems that the issue is OpenCL related.
OtherOS region
OtherOS/OtherOS++ region is on HDD (ps3dd), we have new linux tools (ps3sed) and drivers.
To resize ps3da I've tried new ps3sed (manually), unsuccesfully: GameOS always detect corruption and redo its own things.
I've found a way to force resize on 4.46, no emer_init patch, no downgrading: GameOS respect standards.
I can now resize ps3da at arbitrary size.
Swapping HDD on pc is necessary to me to send a couple to SET MAX ADDRESS ata commands to get the job done: set the limit, left GameOS (partition and) format, then reset size back the same way.
On boot all regions are fine, plus empty space as tail, nice to fit a fouth region.
Here I've forced ps3da to use 1216709344 sectors, this left me about 16G for ps3dd.
After that GameOS do it own things, I've resetted ps3da to its real geometry (1250263728) and booted a new petitboot.
root@ps3-linux:~# dmesg | grep ps3disk [ 3.220526] ps3disk_init:601: registered block device major 254 [ 3.220549] ps3_system_bus_match:369: dev=6.0(sb_04), drv=6.0(ps3disk): match [ 3.220856] ps3disk sb_04: accessible region 0 start 0 size 1250263728 [ 3.220952] ps3disk sb_04: accessible region 1 start 32 size 1212515008 [ 3.221045] ps3disk sb_04: accessible region 2 start 1212515040 size 4194296 [ 3.221051] ps3disk sb_04: ps3stor_probe_access:133: 3 accessible regions found [ 3.227341] ps3disk sb_04: ps3da is a SAMSUNG HM641JI (610480 MiB total, 610480 MiB region) [ 3.229035] ps3disk sb_04: ps3db is a SAMSUNG HM641JI (610480 MiB total, 592048 MiB region) [ 3.230008] ps3disk sb_04: ps3dc is a SAMSUNG HM641JI (610480 MiB total, 2047 MiB region) root@ps3-linux:~# ps3sed print_region 3 0 0 1250263728 1 1 32 1212515008 8 2 1212515040 4194296 8 root@ps3-linux:~# create_hdd_region.sh INFO: device id 3 INFO: number of regions 3 INFO: total number of blocks 1250263728 INFO: last region start block 1212515040 INFO: last region number of blocks 4194296 INFO: new region start block 1216709344 INFO: new region number of blocks 33554376 INFO: new region id 3 root@ps3-linux:~# ps3sed print_region 3 0 0 1250263728 1 1 32 1212515008 8 2 1212515040 4194296 8 3 1216709344 33554376 1 root@ps3-linux:~# reboot && exit
Last number 1 is wrong, it says that last region has only one acl entry, we need to fix it at 8 entries:
- manually with ps3sed
- rebooting
Petitboot finally detect a new ps3dd device, the fourth region, of (33554376 * 512 =) 17179840512 bytes.
All of this with a 3.10.26 kernel and new tools: no vflash hacking involved, linux on vflash7 is deprecated.
Sometimes HDD is reported as second device (something buggy in my kernel?):
root@ps3-linux:~# ps3sed print_device
flash 1 512 491008 7
cdrom 3 2048 2147483647 1
disk 2 512 1250263728 4
00000000 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
00000010 00 00 00 00 0f ac e0 ff 00 00 00 00 de ad fa ce |................|
00000020 00 00 00 00 00 00 00 03 00 00 00 00 00 00 00 02 |................|
00000030 00 00 00 00 00 00 00 20 00 00 00 00 48 45 82 c0 |....... ....HE..|
00000040 10 70 00 00 02 00 00 01 00 00 00 00 00 00 00 03 |.p..............|
00000050 10 70 00 00 01 00 00 01 00 00 00 00 00 00 00 03 |.p..............|
00000060 10 20 00 00 03 00 00 01 00 00 00 00 00 00 00 03 |. ..............|
00000070 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
*
000000c0 00 00 00 00 48 45 82 e0 00 00 00 00 00 3f ff f8 |....HE.......?..|
000000d0 10 70 00 00 02 00 00 01 00 00 00 00 00 00 00 03 |.p..............|
000000e0 10 70 00 00 01 00 00 01 00 00 00 00 00 00 00 03 |.p..............|
000000f0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
*
00000150 00 00 00 00 48 85 82 e0 00 00 00 00 01 ff ff c8 |....H...........|
00000160 10 70 00 00 02 00 00 01 00 00 00 00 00 00 00 03 |.p..............|
00000170 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
*
00000400
ps3vuart-tools
We miss some stuff from old ps3sm-utils, looking to port: temperature, get_fan_policy and set_fan_policy to new ps3vuart-tools.
We need to enable some sort of fan control on petitboot now.
root@fedora_clone ~]# /home/ps3vuart-tools-2012-09-01/ps3sm/ps3sm get_fan_policy 0 0x01 0x01 0x48 0x00 [root@fedora_clone ~]# /home/ps3vuart-tools-2012-09-01/ps3sm/ps3sm temperature 0 01 00 00 00 3f 49 00 00
Updating the Real Time Clock with hwclock results in error:
Mar 31 18:09:18 fedora_clone kernel: os_area_queue_work_handler: Could not update FLASH ROM
UPL.xml.pkg
tar -t -f update_files.tar ls UPL.xml.unpkg/ -rw-r--r-- 1 0 0 2.8K Jun 27 13:40 content -rw-r--r-- 1 0 0 64 Jun 27 13:40 info0 -rw-r--r-- 1 0 0 64 Jun 27 13:40 info1 ... -rwxr-xr-x 1 0 0 640 Jun 27 15:20 UPL.xml.pkg.spkg_hdr.1 UPL.xml.unpkg/content: XML document text UPL.xml.unpkg/info0: data UPL.xml.unpkg/info1: data UPL.xml.unpkg/UPL.xml.pkg.spkg_hdr.1: data
scetool -v -i update_files.untar/UPL.xml.pkg [*] Using keyset [pkg 0x0000 03.55] [*] Header decrypted. [*] Data decrypted. [*] SCE Header: Magic 0x53434500 [OK] Version 0x00000002 Key Revision 0x0000 Header Type [PKG] Metadata Offset 0x00000000 Header Length 0x0000000000000280 Data Length 0x0000000000000B9D // 2973 bytes, content + info0 + info1 [*] Metadata Info: Key 87 EE 46 44 60 DA DA EA 49 74 58 F9 02 1D 6D 11 IV F4 9F 43 D8 D0 6A F0 FC 33 AF 5E 6E CF 2F 30 1E [*] Metadata Header: Signature Input Length 0x0000000000000250 unknown_0 0x00000001 Section Count 0x00000003 Key Count 0x00000014 Optional Header Size 0x00000000 unknown_1 0x00000000 unknown_2 0x00000000 [*] Metadata Section Headers: Idx Offset Size Type Index Hashed SHA1 Encrypted Key IV Compressed 000 00000280 00000040 01 01 [YES] 00 [NO ] -- -- [NO ] 001 000002C0 00000040 02 02 [YES] 06 [NO ] -- -- [NO ] 002 00000300 0000016B 03 03 [YES] 0C [YES] 12 13 [YES] [*] SCE File Keys: n 14
hexdump -C UPL.xml.pkg
00000280 00 00 00 03 00 00 00 04 00 00 00 00 00 00 00 0a |................| 00000290 20 14 06 19 01 15 45 00 00 00 00 00 00 00 0b 1d | .....E.........| 000002a0 00 00 00 00 00 00 01 6b 00 00 00 00 00 00 00 00 |.......k........| 000002b0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................| 000002c0 00 00 00 00 00 00 00 03 00 00 00 00 00 00 00 40 |...............@| 000002d0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0b 1d |................| 000002e0 00 00 00 00 00 00 00 01 00 00 00 00 00 00 00 01 |................| 000002f0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................| We got info0 + info1, 64bytes each content is 2845bytes, total is data lenght 2973 info0, info1 are not encrypted
UPL.xml.pkg is decrypted using pkg keyset [0x0000 03.55], UPL.xml.pkg.spkg_hdr.1 is decrypted using spkg keyset for > 3.55: only Metadata Info and Signature changes.
I think that on 3.55 all .spkg_hdr.1 files are not involved in updating, on 3.56 all of them are used as in overlayfs, so all .spkg_hdr.1 files are readed as metadata headers, decrypted by newer spkg keyset to get the same SCE keys to get rest of data decrypted.
- https://www.certicom.com/index.php/20-elliptic-curve-groups-over-real-numbers
- http://csrc.nist.gov/groups/ST/key_mgmt/documents/June09_Presentations/rene.struik.KMWJune09_5Min.pdf
- http://cacr.uwaterloo.ca/techreports/2005/cacr2005-28.pdf
- http://stackoverflow.com/questions/20266368/ecdsa-signature-size-too-long
.spkg_hdr.1
strip metadata info
# dd if=CORE_OS_PACKAGE.pkg.spkg_hdr.1 of=metai.crypt skip=32 count=64 bs=1
64+0 records in
64+0 records out
64 bytes (64 B) copied, 0.0021009 s, 30.5 kB/s
# hexdump -C metai.cry
00000000 d7 f9 82 9e 75 0a 3f 20 8b 6f e7 41 b1 bb 52 15 |....u.? .o.A..R.|
00000010 e1 8f d2 86 43 b5 4f 56 4c 42 a0 10 e1 1a 25 38 |....C.OVLB....%8|
00000020 9c 28 c7 fd 38 31 24 3b 1b 2b 9f 3f dc 72 4f c4 |.(..81$;.+.?.rO.|
00000030 95 34 b8 0a af 25 a1 05 b6 8f ce 2c 88 e9 2b 7b |.4...%.....,..+{|
verify metadata info decryption with standard tool
# openssl enc -d -aes-256-cbc -in metai.crypt -K erk -iv riv | hexdump -C 00000000 7c f2 9a 4b 96 de 5f 75 a1 32 87 c0 42 ec 8f cf ||..K.._u.2..B...| 00000010 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................| 00000020 6f 85 6a 60 2a 8d b4 3f 2a 81 1b 1a 9c a3 02 f6 |o.j`*..?*.......| 00000030
decrypt rest of metadata
get signature
compute digest of the whole metadata
verify digest / signature