aboutsummaryrefslogtreecommitdiffstats
path: root/fs
diff options
context:
space:
mode:
authorKeith Mannthey <kmannth@us.ibm.com>2008-11-25 04:24:35 -0500
committerJens Axboe <jens.axboe@oracle.com>2008-12-29 02:28:44 -0500
commit08bafc0341f2f7920e9045bc32c40299cac8c21b (patch)
treeab4ad0f73f457409239a98c050bf7edd47c03cf1 /fs
parent7c239517d9f18427fc2e7ed259fb3b866595f5af (diff)
block: Supress Buffer I/O errors when SCSI REQ_QUIET flag set
Allow the scsi request REQ_QUIET flag to be propagated to the buffer file system layer. The basic ideas is to pass the flag from the scsi request to the bio (block IO) and then to the buffer layer. The buffer layer can then suppress needless printks. This patch declutters the kernel log by removed the 40-50 (per lun) buffer io error messages seen during a boot in my multipath setup . It is a good chance any real errors will be missed in the "noise" it the logs without this patch. During boot I see blocks of messages like " __ratelimit: 211 callbacks suppressed Buffer I/O error on device sdm, logical block 5242879 Buffer I/O error on device sdm, logical block 5242879 Buffer I/O error on device sdm, logical block 5242847 Buffer I/O error on device sdm, logical block 1 Buffer I/O error on device sdm, logical block 5242878 Buffer I/O error on device sdm, logical block 5242879 Buffer I/O error on device sdm, logical block 5242879 Buffer I/O error on device sdm, logical block 5242879 Buffer I/O error on device sdm, logical block 5242879 Buffer I/O error on device sdm, logical block 5242872 " in my logs. My disk environment is multipath fiber channel using the SCSI_DH_RDAC code and multipathd. This topology includes an "active" and "ghost" path for each lun. IO's to the "ghost" path will never complete and the SCSI layer, via the scsi device handler rdac code, quick returns the IOs to theses paths and sets the REQ_QUIET scsi flag to suppress the scsi layer messages. I am wanting to extend the QUIET behavior to include the buffer file system layer to deal with these errors as well. I have been running this patch for a while now on several boxes without issue. A few runs of bonnie++ show no noticeable difference in performance in my setup. Thanks for John Stultz for the quiet_error finalization. Submitted-by: Keith Mannthey <kmannth@us.ibm.com> Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Diffstat (limited to 'fs')
-rw-r--r--fs/buffer.c19
1 files changed, 15 insertions, 4 deletions
diff --git a/fs/buffer.c b/fs/buffer.c
index 10179cfa1152..776ae091d3b0 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -99,10 +99,18 @@ __clear_page_buffers(struct page *page)
99 page_cache_release(page); 99 page_cache_release(page);
100} 100}
101 101
102
103static int quiet_error(struct buffer_head *bh)
104{
105 if (!test_bit(BH_Quiet, &bh->b_state) && printk_ratelimit())
106 return 0;
107 return 1;
108}
109
110
102static void buffer_io_error(struct buffer_head *bh) 111static void buffer_io_error(struct buffer_head *bh)
103{ 112{
104 char b[BDEVNAME_SIZE]; 113 char b[BDEVNAME_SIZE];
105
106 printk(KERN_ERR "Buffer I/O error on device %s, logical block %Lu\n", 114 printk(KERN_ERR "Buffer I/O error on device %s, logical block %Lu\n",
107 bdevname(bh->b_bdev, b), 115 bdevname(bh->b_bdev, b),
108 (unsigned long long)bh->b_blocknr); 116 (unsigned long long)bh->b_blocknr);
@@ -144,7 +152,7 @@ void end_buffer_write_sync(struct buffer_head *bh, int uptodate)
144 if (uptodate) { 152 if (uptodate) {
145 set_buffer_uptodate(bh); 153 set_buffer_uptodate(bh);
146 } else { 154 } else {
147 if (!buffer_eopnotsupp(bh) && printk_ratelimit()) { 155 if (!buffer_eopnotsupp(bh) && !quiet_error(bh)) {
148 buffer_io_error(bh); 156 buffer_io_error(bh);
149 printk(KERN_WARNING "lost page write due to " 157 printk(KERN_WARNING "lost page write due to "
150 "I/O error on %s\n", 158 "I/O error on %s\n",
@@ -394,7 +402,7 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate)
394 set_buffer_uptodate(bh); 402 set_buffer_uptodate(bh);
395 } else { 403 } else {
396 clear_buffer_uptodate(bh); 404 clear_buffer_uptodate(bh);
397 if (printk_ratelimit()) 405 if (!quiet_error(bh))
398 buffer_io_error(bh); 406 buffer_io_error(bh);
399 SetPageError(page); 407 SetPageError(page);
400 } 408 }
@@ -455,7 +463,7 @@ static void end_buffer_async_write(struct buffer_head *bh, int uptodate)
455 if (uptodate) { 463 if (uptodate) {
456 set_buffer_uptodate(bh); 464 set_buffer_uptodate(bh);
457 } else { 465 } else {
458 if (printk_ratelimit()) { 466 if (!quiet_error(bh)) {
459 buffer_io_error(bh); 467 buffer_io_error(bh);
460 printk(KERN_WARNING "lost page write due to " 468 printk(KERN_WARNING "lost page write due to "
461 "I/O error on %s\n", 469 "I/O error on %s\n",
@@ -2913,6 +2921,9 @@ static void end_bio_bh_io_sync(struct bio *bio, int err)
2913 set_bit(BH_Eopnotsupp, &bh->b_state); 2921 set_bit(BH_Eopnotsupp, &bh->b_state);
2914 } 2922 }
2915 2923
2924 if (unlikely (test_bit(BIO_QUIET,&bio->bi_flags)))
2925 set_bit(BH_Quiet, &bh->b_state);
2926
2916 bh->b_end_io(bh, test_bit(BIO_UPTODATE, &bio->bi_flags)); 2927 bh->b_end_io(bh, test_bit(BIO_UPTODATE, &bio->bi_flags));
2917 bio_put(bio); 2928 bio_put(bio);
2918} 2929}