SDHC kernel updated - mk3 - maemo.org

Active Topics

Camera phone competition March 2023 "Endless" (35)
to General by Maemish - 1 day, 14 hrs ago
TLS 1.2 for N9 is ready (28)
to MeeGo / Harmattan by smartblu9 - 1 day, 15 hrs ago
Maemo repository (3)
to Community by torsen - 2 days, 7 hrs ago
Error after installing Leste on sdcard (11)
to Maemo 7 / Leste by Maemish - 3 days, 5 hrs ago
Introducing ubiboot N9 (multiboot OS loader) (1,393)
to Alternatives by smatkovi - 4 days, 11 hrs ago
more...

Page 1 of 2

Next >

Thread Tools

Milhouse	2007-03-11 , 19:20
Posts: 3,401 \| Thanked: 1,255 times \| Joined on Nov 2005 @ London, UK	#1

http://intr.overt.org/blog/?p=49

2.2006.51-6 kernel updated with the latest SDHC/MMC patches, fixing low-voltage MMC and high-speed MMC issues.

Quote & Reply |

fanoush	2007-03-12 , 12:53
Posts: 2,152 \| Thanked: 1,490 times \| Joined on Jan 2006 @ Czech Republic	#2

Anyone with Transcend 150x 4GB (non-SDHC one) could post clock debug log from this SDHC kernel and also test the speed like this (two times, post the second/slower one)?

Code:

Nokia-N800-51:~# time dd if=/dev/mmcblk1 of=/dev/null bs=8192 count=30000

I've got older 80x transcend 1GB which does run in high-speed SD mode (i.e. 48MHz) but still the read speed is ~7MB/s. I wonder whether the 150x one could be better (it should).

The kingston 2GB mmcmobile is my best card so far, it runs at 48Mhz over 4bits and read speed in N800 is 12MB/s, which is great. Can the 4GB 150x one match or exceed this?

Code:

Nokia-N800-51:~# time dd if=/dev/mmcblk1 of=/dev/null bs=8192 count=30000
30000+0 records in
30000+0 records out
real    0m 18.76s
user    0m 0.03s
sys     0m 3.07s
Nokia-N800-51:~#

240MB in 19 seconds =~ 12.5 MB/s

Quote & Reply |

Milhouse	2007-03-12 , 17:30
Posts: 3,401 \| Thanked: 1,255 times \| Joined on Nov 2005 @ London, UK	#3

Fanoush - I've got similar results when comparing high-speed and low-speed cards, though I'm not sure why there is such a difference between the 8GB and 4GB Transcend cards, they're seemingly identical other than capacity however the higher capacity card is significantly faster on reads.

Code:

Test: time dd if=/dev/mmcblk? of=/dev/null bs=8192 count=30000
Speed in bytes => 8192*30000/time
Speed in Mbytes => 8192*30000/time/1024/1024

Card	               Clock    Width     Time       Speed        Slot
1GB SanDisk RS-MMC     20Mhz   1-bits   107.29    2.18MB/s    External
4GB SanDisk SDHC C2    25Mhz   4-bits    31.41    7.46MB/s    External
4GB Transcend SDHC C2  48Mhz   4-bits    25.72    9.11MB/s    Internal
8GB Transcend SDHC C2  48Mhz   4-bits    19.54   11.99MB/s    Internal

Last edited by Milhouse; 2007-03-12 at 17:46.

Quote & Reply |

digitalanalogue	2007-03-16 , 10:31
Posts: 7 \| Thanked: 0 times \| Joined on Mar 2007	#4

I flashed my n800 with this kernel.bin last night and lo my 8gb Transcend sdhc card works!

Now I'm at work listening to some of the many MP3's I've now got on the card in Canola. Occasionally there's a glitch (slight pause) early on in a track, if I restart the track I don't get the glitch, so its not in the file itself (happens on several different tracks too). I also never had this happen with my other 2gb stard sd card.

I'm just wondering what could be causing this, the README on Philip's webpage (dir with the bin file) mentions something about a debug file being written, could this be happening?

I'm not fully up to speed with linux yet, so I'm struggling to find said debug file, otherwise I could just check the time/date on it.

Any advice would be appreciated.

Quote & Reply |

CliffOn	2007-03-16 , 10:33
Posts: 8 \| Thanked: 0 times \| Joined on Mar 2007 @ SoCal	#5

I don't know how to get the debug logs.
Here is the timing for my two cards mmcblk0 being the Transcend 150x 2gig (internal) and mmcblk1 being a not even specced (none included in blister pack) Kingston 2gig. I tested before installing the sdhc kernel and the mmcblk0 was 32 secs and the kingston 34 secs, didn't screen cap it :/

the results below are unexpected, I expected the transcend to be in line with the other results here, the kingston I expected nothing from and I got a roughly 1/3 gain.
Cliff

Nokia-N800-51:~# time dd if=/dev/mmcblk0 of=/dev/null bs=8192 count=30000
30000+0 records in
30000+0 records out
real 0m 20.91s
user 0m 0.03s
sys 0m 3.69s
Nokia-N800-51:~# time dd if=/dev/mmcblk0 of=/dev/null bs=8192 count=30000
30000+0 records in
30000+0 records out
real 0m 20.93s
user 0m 0.04s
sys 0m 3.78s
Nokia-N800-51:~# time dd if=/dev/mmcblk1 of=/dev/null bs=8192 count=30000
30000+0 records in
30000+0 records out
real 0m 21.59s
user 0m 0.07s
sys 0m 3.88s

__________________
N800
Sierra Stowaway Full-sized KB. If you can't stand an embedded number pad, get this, 80 bux at amazon.
Rootfs running off 2 gig transcend 150X SD
Cellink BT GPS

Quote & Reply |

fanoush	2007-03-16 , 11:33
Posts: 2,152 \| Thanked: 1,490 times \| Joined on Jan 2006 @ Czech Republic	#6

kernel log is recorded to kernel buffer in RAM so it should not slow down anything. This one adds just few lines about MMC clock speed when card is inserted. You can print it via dmesg command typed in osso-xterm. As the buffer is circular and limited by size you see only last messages.

Thanks CliffOn for your numbers, looks like the 150x one is good choice (speedwise).

Quote & Reply |

The Following User Says Thank You to fanoush For This Useful Post:

digitalanalogue	2007-03-16 , 11:47
Posts: 7 \| Thanked: 0 times \| Joined on Mar 2007	#7

I'll try the previous SDHC kernel, maybe the latest is too optimised for speed for my card?

Quote & Reply |

rok	2007-03-17 , 11:06
Posts: 66 \| Thanked: 7 times \| Joined on Nov 2005 @ Berlin, Germany	#8

I'm experienceing the same glitches with my 8gb Transcend while listening to my mp3s.
I'm still using the first SDHC kernel.

Anyone else?

__________________
greetings
rok

Beware of my English! Beware of my English!

Quote & Reply |

digitalanalogue	2007-03-20 , 13:59
Posts: 7 \| Thanked: 0 times \| Joined on Mar 2007	#9

Just realised these glitches only happen in Canola, so I can live with using the nokia media player, still pretty happy with having a 10gb media player / internet browser!

Rok, do you have the same results with media player?

Quote & Reply |

vash	2007-04-14 , 00:36
Posts: 19 \| Thanked: 0 times \| Joined on Mar 2007 @ Ontario, Canada	#10

Test Results for 2 cards: OCZ 1 GB 133X and ATP ProMAX 4GB SDHC Class 6

Card placement: the ATP is in the Internal Bay, OCZ is in the external slot.

Results:

Code:

Nokia-N800-10:~# time dd if=/dev/mmcblk1 of=/dev/null bs=8192 count=30000
30000+0 records in
30000+0 records out
real    0m 20.31s
user    0m 0.07s
sys     0m 2.71s
Nokia-N800-10:~# time dd if=/dev/mmcblk1 of=/dev/null bs=8192 count=30000
30000+0 records in
30000+0 records out
real    0m 21.12s
user    0m 0.02s
sys     0m 2.52s
Nokia-N800-10:~# time dd if=/dev/mmcblk0 of=/dev/null bs=8192 count=30000
30000+0 records in
30000+0 records out
real    0m 18.85s
user    0m 0.10s
sys     0m 2.89s
Nokia-N800-10:~# time dd if=/dev/mmcblk0 of=/dev/null bs=8192 count=30000
30000+0 records in
30000+0 records out
real    0m 18.94s
user    0m 0.08s
sys     0m 2.73s

Is there any way we can test write speeds using something similar to this method? That Kingston looks pretty good, considering that this ATP is not exactly cheap.