[maemo-developers] Firmware 4.2007.26-8 has inferior SDHC performance than 3.2007.10-7 - why?

[From: Neil MacLeod]

Frantisek Dufka wrote:
> It may be that MMC speed is limited to 24Mhz (wonder what NB#54313 is 
> about) for similar reason like N770 MMC speed being limited to 12Mhz - 
> better compatibility by sacrificing speed. Sadly this makes perfect 
> sense for Nokia.
> 
I suspected that to be the situation with the clock capped at 24Mhz for compatibility reasons however I had more compatibility and seemingly fewer issues with the 48Mhz full fat SDHC patch in 3.2007.10-8 than I have with 4.2007.26-8. I've now experienced one corrupt (but fortunately not killed) FAT filesystem on a 4GB SDHC Class 2 Transcend card.

In the 3+ months of using the full speed 3.2007.10-8 patch on two N800s I never experienced any corruption yet it has now occurred within 36 hours following the upgrade (and I've only installed the update on one of the N800s!)

I wonder if Nokia have jumped the gun and capped the speed unnecessarily when the corruption issue lies elsewhere? Assuming this to be the case, will we ever get full speed support in future? Judging from the (lack of problem) reports over the last 3+ months the high speed patch didn't seem to cause any major issues, unlike the new firmware - quite a number of reports have appeared on ITT[1,2,3] suggesting corrupt, dead and unrecoverable SDHC cards have resulted from using the latest firmware (and bug has had to be re-opened in bugzilla[4]).

I agree that the ~6MB/s read speed is good, but it's a little annoying that we all know the N800 can achieve at least double this performance and is being artificially restricted for possibly the wrong reasons. :)

1. http://www.internettablettalk.com/forums/showthread.php?t=7384
2. http://www.internettablettalk.com/forums/showthread.php?t=5989
3. http://www.internettablettalk.com/forums/showthread.php?p=57696#post57696
4. https://bugs.maemo.org/show_bug.cgi?id=1204