On Tue, 2010-01-05 at 09:48 -0600, Paul Hartman wrote: > 2010/1/5 Kimmo Hämäläinen <kimmo.hamalainen@xxxxxxxxx>: > > Hi, > > > > On Sun, 2009-12-27 at 22:38 +0100, ext Paul Hartman wrote: > >> Hi, > >> > >> I'm using linux kernel 2.6.32 on my PC and mount N900 in mass storage > >> mode using vfat (async). > >> > >> If I write one file then sync/umount, write speed is about 17MB/sec, > >> same as I get on Windows Vista. However, if I write more than one file > >> and then sync/umount, speed is terrible, around 2MB/sec (and in kernel > >> 2.6.31 that speed was about 450kb/sec using the old pdflush code). So > >> it seems linux is maybe creating multiple write streams and N900 > >> doesn't deal with that well at all. > > > > There has been some USB speed-ups since the sales release, but I'm not > > sure if they fix this multiple file copying problem. I think it's best > > to file a bug to maemo.org, so Maemo kernel guys get the word (I think > > too few of them read this mailing list). > > Done. Will also post to LKML in case it's a general problem with linux > i/o scheduler on the PC side. > > https://bugs.maemo.org/show_bug.cgi?id=7677 I don't think LKML will appreciate debugging a problem in a kernel with closed-source components. Xav _______________________________________________ maemo-users mailing list maemo-users@xxxxxxxxx https://lists.maemo.org/mailman/listinfo/maemo-users
