Hello,
I am new to device-mapper and block layer. In my device mapper driver I use generic_make_request() function to pass the bio to lower layer. But I see this function wait for IO for most of the time that cause my driver throughput be very low. The IOs my driver pass down is all 4K size.
Could someone please help me or point me to some documentation to understand this issue? Shouldn't this function be returning immediately as I am passing the completion function along with the bio ?
Thanks in advance,
-Siju
-- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
