Messages in this thread Patch in this message | | | Subject | Re: Slowwwwwwwwwwww NFS read performance.... | From | Trond Myklebust <> | Date | Thu, 15 Jan 2004 15:54:49 -0500 |
| |
På to , 15/01/2004 klokka 15:16, skreiv Trond Myklebust: > På to , 15/01/2004 klokka 14:53, skreiv Ram Pai: > > Yes this problem has been reported earlier. Attaching Andrew's patch > > that reverts a change. This should work. Please confirm. > > Sorry to disappoint you, but that change appears already in 2.6.1-mm3, > and does not suffice to fix the problem.
The following reversion is what fixes my regression. That puts the sequential read numbers back to the 2.6.0 values of ~140MB/sec (from the current 2.6.1 values of 14MB/second)...
Cheers, Trond
diff -u --recursive --new-file linux-2.6.1-mm3/mm/readahead.c linux-2.6.1-fixed/mm/readahead.c --- linux-2.6.1-mm3/mm/readahead.c 2004-01-09 01:59:07.000000000 -0500 +++ linux-2.6.1-fixed/mm/readahead.c 2004-01-15 15:41:35.118000000 -0500 @@ -474,13 +474,9 @@ /* * This read request is within the current window. It is time * to submit I/O for the ahead window while the application is - * about to step into the ahead window. - * Heuristic: Defer reading the ahead window till we hit - * the last page in the current window. (lazy readahead) - * If we read in earlier we run the risk of wasting - * the ahead window. + * crunching through the current window. */ - if (ra->ahead_start == 0 && offset == (ra->start + ra->size -1)) { + if (ra->ahead_start == 0) { ra->ahead_start = ra->start + ra->size; ra->ahead_size = ra->next_size; actual = do_page_cache_readahead(mapping, filp, - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |