At 08:05 AM 3/19/2003, Callebaut, Beno?t wrote: >After a lot of debugging I seen that there is a bug in the MTD read only >block driver. >It seems that in the mtdblock_request, something goes wrong withe the >handling of the request structure. >between two calls to end_request, the pointer bh->req_qnext of the last used >buffer head is corrupted. >I didn't find the cause but since this happens only with the >mtdblock_request function I suppose there is something wrong in it. > >If I ever find what's wrong I will try to correct it, but don't hope too >much I don't have a lot of time.
I submitted fix for mtdblock_ro.c to MTD CVS several months ago. Please, get the latest mtdblock_ro.c from http://www.linux-mtd.infradead.org/, it should solve your problems. Eugene ** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/
