Jens Axboe [Thu, 2 Feb 2006 12:47:51 +0000 (13:47 +0100)]
[PATCH] blkparse: Fix bad ordering due to non-global sequence
We need to let events simmer in the queue for at least one iteration
of all devices, to prevent bad outputs if one CPU has late events.
Jens Axboe [Thu, 2 Feb 2006 12:20:57 +0000 (13:20 +0100)]
[PATCH] blkparse: make skip detection per-CPU as well
Jens Axboe [Thu, 2 Feb 2006 12:20:03 +0000 (13:20 +0100)]
[PATCH] verify_blkparse: improve output
Jens Axboe [Thu, 2 Feb 2006 11:06:34 +0000 (12:06 +0100)]
[PATCH] verify_blkparse: update for per-CPU sequences
Jens Axboe [Thu, 2 Feb 2006 11:00:14 +0000 (12:00 +0100)]
[PATCH] .gitignore: add blkrawverify
Jens Axboe [Thu, 2 Feb 2006 10:56:56 +0000 (11:56 +0100)]
[PATCH] Make the sequence per-CPU so the kernel side can be lockless
Jens Axboe [Thu, 2 Feb 2006 10:16:05 +0000 (11:16 +0100)]
[PATCH] blktrace: remove check_time, we always need to check it
Jens Axboe [Thu, 2 Feb 2006 08:23:56 +0000 (09:23 +0100)]
[PATCH] blktrace: make write_data() return early with buflen == 0
Alan D. Brunelle [Thu, 2 Feb 2006 08:23:20 +0000 (09:23 +0100)]
[PATCH] blktrace: speedup write out by aggregating writes
Alan D. Brunelle [Thu, 2 Feb 2006 08:16:58 +0000 (09:16 +0100)]
[PATCH] blktrace: remember to initialize mutex
Jens Axboe [Tue, 31 Jan 2006 19:21:15 +0000 (20:21 +0100)]
[PATCH] kernel: update to include Alan's relay_reserve() method
Jens Axboe [Tue, 31 Jan 2006 08:24:29 +0000 (09:24 +0100)]
[PATCH] blkparse: be sure to propagate read_data() error all the way back
Jens Axboe [Tue, 31 Jan 2006 08:21:04 +0000 (09:21 +0100)]
[PATCH] blkparse: cut down on excessive number of fcntl() calls
Cache the blocking state locally and only issue it if changed.
Jens Axboe [Tue, 31 Jan 2006 08:11:01 +0000 (09:11 +0100)]
[PATCH] blkparse: don't do partial reads in read_data()
If we do a partial read, the trace may look ok because the magic is there,
but we will have lost it and the next read() will be confused. So only
return early if we read zero bytes and we must not block, otherwise wait
around for the remaining data.
Jens Axboe [Mon, 30 Jan 2006 20:35:31 +0000 (21:35 +0100)]
[PATCH] btrace: don't be quiet by default, and fix -h parameter
Alan D. Brunelle [Mon, 30 Jan 2006 18:03:43 +0000 (19:03 +0100)]
[PATCH] blkparse: Remove negative skips
The following patch does not insert skips that go backwards. I am not
sure why we do, but on occasion we find events that are behind where we
should be, and we don't have an entry in skipped ones - that would make
me thing it's a duplicate. But in any event, we should not be reporting
such things - they cause huge skip numbers by wrapping on an unsigned long.
Jens Axboe [Mon, 30 Jan 2006 13:37:04 +0000 (14:37 +0100)]
[PATCH] blkparse: don't let SIGINT show stats right away, we might not be done
Jens Axboe [Mon, 30 Jan 2006 13:36:25 +0000 (14:36 +0100)]
[PATCH] blktrace: allow live tracing to get full subbuffers as well
The poll() timeout will keep us reading smaller buffers, if the data
generation rate isn't high enough.
Jens Axboe [Mon, 30 Jan 2006 13:31:01 +0000 (14:31 +0100)]
[PATCH] blktrace: fix leftover_ts offsetting
Jens Axboe [Sat, 28 Jan 2006 19:39:11 +0000 (20:39 +0100)]
[PATCH] blkrawverify: Check for EOF in error reporting
A check of fread returning 0 is not sufficient towards determining an.
error - 0 also indicates EOF. Added check for that.
Jens Axboe [Sat, 28 Jan 2006 19:36:09 +0000 (20:36 +0100)]
[PATCH] blktrace: really fix the loop logic
Jens Axboe [Sat, 28 Jan 2006 19:15:27 +0000 (20:15 +0100)]
[PATCH] blktrace: fix logic for write out loop when exiting
Jens Axboe [Sat, 28 Jan 2006 19:06:28 +0000 (20:06 +0100)]
[PATCH] blkrawverify: cast __u64 to unsigned long long for printing
Jens Axboe [Fri, 27 Jan 2006 18:42:30 +0000 (19:42 +0100)]
[PATCH] blktrace: fix hang and no output if first sequence is not expected
Alan D. Brunelle [Fri, 27 Jan 2006 16:12:35 +0000 (17:12 +0100)]
[PATCH] blktrace: cleanup the leftover subbuf code
Jens Axboe [Fri, 27 Jan 2006 15:26:25 +0000 (16:26 +0100)]
[PATCH] blktrace/blkrawverify: get rid of pesky signedness warnings
Jens Axboe [Fri, 27 Jan 2006 13:47:15 +0000 (14:47 +0100)]
[PATCH] Need to still use outputname to check no_stdout
Jens Axboe [Fri, 27 Jan 2006 13:41:59 +0000 (14:41 +0100)]
[PATCH] Forgot to add list.h
Jens Axboe [Fri, 27 Jan 2006 13:39:35 +0000 (14:39 +0100)]
[PATCH] blktrace: completely rewrite the event pull/pushing loops
Jens Axboe [Fri, 27 Jan 2006 09:35:47 +0000 (10:35 +0100)]
[PATCH] kernel: update kernel patch with the HT sibling time offset fix
Jens Axboe [Thu, 26 Jan 2006 19:07:26 +0000 (20:07 +0100)]
[PATCH] blkrawverify: u64 types need %llu format
Alan D. Brunelle [Thu, 26 Jan 2006 15:03:12 +0000 (16:03 +0100)]
[PATCH] Added blkrawverify utility
Jens Axboe [Thu, 26 Jan 2006 05:48:45 +0000 (21:48 -0800)]
[PATCH] blktrace: improve live tracing
Jens Axboe [Wed, 25 Jan 2006 17:59:38 +0000 (18:59 +0100)]
[PATCH] kernel: update patch for the SMP fix
Jens Axboe [Wed, 25 Jan 2006 17:54:14 +0000 (18:54 +0100)]
[PATCH] blkparse: fill unitialized pci->cpus.fd with -1
Tom Zanussi [Wed, 25 Jan 2006 14:10:43 +0000 (15:10 +0100)]
[PATCH] blktrace: Don't overwrite buf_size for the vbuffer
Here's another patch to fix a problem I noticed - the code that sets
the buffer size passed to setvbuf() uses the global buf_size variable
and thus unintentionally sets that variable, which is used elsewhere
later e.g. setting tip->fd_max_size in extract().
root [Wed, 25 Jan 2006 23:29:47 +0000 (15:29 -0800)]
Merge branch 'master' of ssh://axboe@router/data/git/blktrace
Jens Axboe [Wed, 25 Jan 2006 23:27:28 +0000 (15:27 -0800)]
[PATCH] blkparse: make sure new pci infos have ->fd initialized to -1
This avoids hangs in blktrace when seeing potentially invalid samples, since
it will attempt to read from fd == 0 (stdin) in that case.
Jens Axboe [Wed, 25 Jan 2006 23:05:58 +0000 (15:05 -0800)]
[PATCH] blkparse: add max read/write queueing depth stats
Jens Axboe [Wed, 25 Jan 2006 13:51:06 +0000 (14:51 +0100)]
[PATCH] kernel: Update patch to include ISSUE fix
Jens Axboe [Tue, 24 Jan 2006 18:08:14 +0000 (19:08 +0100)]
Merge branch 'master' of ssh://axboe@router.home.kernel.dk/data/git/blktrace
Alan D. Brunelle [Tue, 24 Jan 2006 18:08:05 +0000 (19:08 +0100)]
[PATCH] blktrace: fix for potential data overwrite
Here is the situation before the realloc:
fd_max_size
|
v
XXXOOOOXXXXX
^
|
fd_off
The X's are data, the O's are empty. What we start with after the realloc:
fd_max_size
|
v
XXXOOOOXXXXXOOOOOOOOOOOO
^
|
fd_off
In the original code we have the movement of data from byte 0
(tip->fd_buf) to tip->fd_off - over-writing data. I think if we use
tip->fd_max_size, we get the desired result (note: tip->fd_max_size is
updated to 2 times the size AFTER the memmove).
Jens Axboe [Sun, 22 Jan 2006 18:15:39 +0000 (19:15 +0100)]
[PATCH] blkparse: hash-by-name is -h, not -n. correct help info.
Tom Zanussi [Sun, 22 Jan 2006 18:10:04 +0000 (19:10 +0100)]
[PATCH] blktrace: bad ret = len assignment that should be ret == len
Alan D. Brunelle [Thu, 19 Jan 2006 14:15:45 +0000 (15:15 +0100)]
[PATCH] blkparse: reverse the order of the skips out for when verbose is enabled
Jens Axboe [Thu, 19 Jan 2006 07:45:04 +0000 (08:45 +0100)]
[PATCH] verify_blkparse: init last_seq to -1 to avoid alias on sequence 0
Jens Axboe [Wed, 18 Jan 2006 19:31:43 +0000 (20:31 +0100)]
[PATCH] blktrace: woops, left debug fprintf() in there, kill it
Jens Axboe [Wed, 18 Jan 2006 19:29:14 +0000 (20:29 +0100)]
[PATCH] blktrace: better poll fix, just return every x msec instead
Jens Axboe [Wed, 18 Jan 2006 19:27:10 +0000 (20:27 +0100)]
[PATCH] blktrace: improve poll() handling, relayfs uses subbuffer granularity
Jens Axboe [Wed, 18 Jan 2006 15:25:16 +0000 (16:25 +0100)]
[PATCH] blktrace: __refile_ringbuffer() needs to return size
Jens Axboe [Wed, 18 Jan 2006 14:00:05 +0000 (15:00 +0100)]
[PATCH] blktrace: signedness fixes
Jens Axboe [Wed, 18 Jan 2006 13:53:23 +0000 (14:53 +0100)]
[PATCH] blktrace: remember to resize ringbuffer before each fill
Jens Axboe [Wed, 18 Jan 2006 13:34:54 +0000 (14:34 +0100)]
[PATCH] blktrace: fix ringbuffer and event drop reaping
Alan D. Brunelle [Wed, 18 Jan 2006 13:33:24 +0000 (14:33 +0100)]
[PATCH] blkparse: improve skip detection / logging
Jens Axboe [Tue, 17 Jan 2006 13:06:44 +0000 (14:06 +0100)]
[PATCH] blktrace: we don't need ->closed anymore
Jens Axboe [Tue, 17 Jan 2006 12:39:00 +0000 (13:39 +0100)]
[PATCH] blktrace: let the threads shut themselves down
This gets rid of the problem where the main and worker threads race
for killing the various file descriptors and buffers.
Jens Axboe [Tue, 17 Jan 2006 12:38:17 +0000 (13:38 +0100)]
[PATCH] blktrace: add ringbuffer in front of kernel buffers
This greatly improves the skipping, I now get full skip free down to 32k
kernel side buffers where I needed 1024k before.
Jens Axboe [Mon, 16 Jan 2006 14:42:58 +0000 (15:42 +0100)]
Merge branch 'master' of ssh://router/data/git/blktrace
Jens Axboe [Mon, 16 Jan 2006 14:42:48 +0000 (15:42 +0100)]
[PATCH] blktrace: kill unused tip members
Jens Axboe [Fri, 13 Jan 2006 14:33:23 +0000 (15:33 +0100)]
[PATCH] kernel: dropped events and cciss support
Jens Axboe [Fri, 13 Jan 2006 14:28:01 +0000 (15:28 +0100)]
[PATCH] blktrace: add warning about increasing buffer size with dropped events
Alan D. Brunelle [Fri, 13 Jan 2006 14:21:00 +0000 (15:21 +0100)]
[PATCH] blkparse: Count forward and backwards skips seperately
Jens Axboe [Fri, 13 Jan 2006 14:05:30 +0000 (15:05 +0100)]
[PATCH] blkparse: dump requeue info as well
Jens Axboe [Fri, 13 Jan 2006 12:26:27 +0000 (13:26 +0100)]
[PATCH] blkparse: __account_c() needs to take a signed integer
Otherwise requeues passing negative byte count don't work as expected.
Jens Axboe [Fri, 13 Jan 2006 08:09:59 +0000 (09:09 +0100)]
[PATCH] blktrace: show stats on sigint as well
Tom Zanussi [Fri, 13 Jan 2006 08:01:15 +0000 (09:01 +0100)]
[PATCH] blktrace: add support for checking events dropped by the kernel
Jens Axboe [Thu, 12 Jan 2006 17:58:33 +0000 (18:58 +0100)]
Merge branch 'master' of ssh://router/data/git/blktrace
Jens Axboe [Thu, 12 Jan 2006 17:57:26 +0000 (18:57 +0100)]
[PATCH] blktrace: smaller code cleanups
Jens Axboe [Thu, 12 Jan 2006 14:05:31 +0000 (15:05 +0100)]
[PATCH] kernel: update kernel patch (mutex update, and device name fix)
This makes blktrace work with device names that includes slashes, such as
cciss (/dev/cciss/c0d0, for instance).
Jens Axboe [Thu, 12 Jan 2006 13:01:14 +0000 (14:01 +0100)]
[PATCH] blktrace: use buffered output, halves the CPU usage of tracing.
Jens Axboe [Thu, 12 Jan 2006 11:44:05 +0000 (12:44 +0100)]
[PATCH] blkparse_fmt: missing 'requeue' event handling
Jens Axboe [Thu, 12 Jan 2006 10:19:24 +0000 (11:19 +0100)]
[PATCH] Fix typo in blk_add_request in elevator.c (kernel patch)
Jens Axboe [Wed, 11 Jan 2006 15:21:34 +0000 (16:21 +0100)]
[PATCH] Increase blktrace default subbuffer size
We generate so much data, that we cannot keep up with the smaller 128kb
buffer. So increase it to 512kb.
Jens Axboe [Wed, 11 Jan 2006 08:47:03 +0000 (09:47 +0100)]
[PATCH] Update kernel patch to newest 2.6.15-git
Jens Axboe [Tue, 10 Jan 2006 20:51:34 +0000 (21:51 +0100)]
[PATCH] Work-around for 32-bit be64_to_cpu() weirdness
Jens Axboe [Wed, 16 Nov 2005 13:07:37 +0000 (14:07 +0100)]
[PATCH] Update kernel patch to match the one in the block git repo
Christoph Hellwig [Wed, 16 Nov 2005 13:04:26 +0000 (14:04 +0100)]
[PATCH] blk_io_trace_remap structure needs padding on 32-bit
Signed-off-by: Jens Axboe <axboe@suse.de>
Jens Axboe [Mon, 31 Oct 2005 10:12:23 +0000 (11:12 +0100)]
[PATCH] Add .gitignore fil
Jens Axboe [Mon, 31 Oct 2005 10:11:31 +0000 (11:11 +0100)]
[PATCH] update kernel support for recent kernel
Also add ->start_lba ->end_lba support for tracing a range, and
->pid for only following a specific app.
Jens Axboe [Tue, 11 Oct 2005 11:14:56 +0000 (13:14 +0200)]
[PATCH] rpm should work for version 3.x now (with changed RPMBUILD)
Jens Axboe [Tue, 11 Oct 2005 11:11:02 +0000 (13:11 +0200)]
[PATCH] Include tar ball in clean target
Jens Axboe [Tue, 11 Oct 2005 11:09:23 +0000 (13:09 +0200)]
[PATCH] Add make rpm target
Jens Axboe [Tue, 11 Oct 2005 10:57:36 +0000 (12:57 +0200)]
[PATCH] 1.0 preparations
Jens Axboe [Tue, 11 Oct 2005 09:36:52 +0000 (11:36 +0200)]
[PATCH] Bump version, should be about good to go for 1.0
Jens Axboe [Tue, 11 Oct 2005 07:50:47 +0000 (09:50 +0200)]
[PATCH] blkparse: be a little more flexible in mixing -i and -D
Alan D. Brunelle [Tue, 11 Oct 2005 07:05:58 +0000 (09:05 +0200)]
[PATCH] kernel: update remap to work for dm as well
Jens Axboe [Mon, 10 Oct 2005 14:54:19 +0000 (16:54 +0200)]
[PATCH] btrace.spec: spelling error
Jens Axboe [Mon, 10 Oct 2005 14:52:51 +0000 (16:52 +0200)]
[PATCH] Add rpm spec file
Jens Axboe [Mon, 10 Oct 2005 14:52:38 +0000 (16:52 +0200)]
[PATCH] blkparse: sparse and friends don't like ULONG_LONG_MAX
Jens Axboe [Mon, 10 Oct 2005 11:28:13 +0000 (13:28 +0200)]
[PATCH] blkparse: fix memory leaks
Jens Axboe [Mon, 10 Oct 2005 10:54:54 +0000 (12:54 +0200)]
[PATCH] blkparse: suff[] should not be static
Jens Axboe [Mon, 10 Oct 2005 10:52:39 +0000 (12:52 +0200)]
[PATCH] blkparse: remember to bail out if conversion gets too large
Jens Axboe [Mon, 10 Oct 2005 10:47:47 +0000 (12:47 +0200)]
[PATCH] blkparse: pretty up stats print by converting to correct K/M/G
Jens Axboe [Mon, 10 Oct 2005 10:20:29 +0000 (12:20 +0200)]
[PATCH] blkparse: kill debug statement in rate calculation
Jens Axboe [Mon, 10 Oct 2005 10:11:16 +0000 (12:11 +0200)]
[PATCH] blkparse: fix bad comment in recent checkin
Jens Axboe [Mon, 10 Oct 2005 10:10:04 +0000 (12:10 +0200)]
[PATCH] blkparse: include rate dump for stats
Jens Axboe [Mon, 10 Oct 2005 10:09:20 +0000 (12:09 +0200)]
[PATCH] blkparse: maintain map of online CPUs instead of open files
The ->nopenfiles fix is not enough, since it doesn't tell us if
the right CPUs are online or not.
Jens Axboe [Sun, 9 Oct 2005 14:07:14 +0000 (16:07 +0200)]
[PATCH] Add large file support for 32-bit platforms
Jens Axboe [Sun, 9 Oct 2005 10:35:49 +0000 (12:35 +0200)]
[PATCH[ verify_blkparse: dump number of events scanned
Jens Axboe [Sun, 9 Oct 2005 10:31:30 +0000 (12:31 +0200)]
[PATCH] stopwatch usage may misorder first event
If the stopwatch_start happens to trigger when we only have
events from _some_ of the available cpus in the rb_sort tree,
the first event may be misreported. Check for that.