aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAge
...
* trace-cmd: Error on record if plugin given does not existSteven Rostedt2011-03-18
| | | | | | | | | If a bad event is given, trace-cmd record fails. But if a bad plugin name is given it continues and does nothing. If a bad plugin name is given it should fail and report an error. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd-record: Add missing put of path name in set_buffer_size()Steven Rostedt2011-03-18
| | | | Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd-record: Have read_file() only update size if givenSteven Rostedt2011-03-18
| | | | Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd-record: Have read_file() NULL terminate the bufferSteven Rostedt2011-03-18
| | | | | | | | Add the '\0' string terminator to the buffer that read the file to make it easier for users of read_file() to treat the read file as a string and not worry about the size. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Enable recording of all CPUs instead of just online onesSteven Rostedt2011-03-18
| | | | | | | | | | | | | | Count all possible CPUs for recording instead of online ones. If we do just online ones, we may not record the ones that are online. We currently just count the CPUs that are on line. If we take CPU 1 offline leaving 3 CPUs online, we record incorrectly CPUs 0-2 instead of CPUs 0,2,3. There may also be a case that we want to record taking a CPU offline or bringing one back on. Reported-by: Steve Clark <Steve.Clark@ffei.co.uk> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Fix NULL pointer deference for use without pluginsSteven Rostedt2011-03-15
| | | | | | | The code that added --func-stack introduced a reference to the variable "plugins" when it was never set to anything. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Add --func-stack option to enable function stack traceSteven Rostedt2011-03-11
| | | | | | | | | | | | | | | | | | Added --func-stack option to trace-cmd record, that will enable the function stack tracing, but only if the -l option (limit functions) is used, and succeeds in filtering functions. It is known that running the function stack trace on all functions can livelock the machine. The output now looks like this: <idle>-0 [002] 64003.696572: function: schedule <idle>-0 [002] 64003.696577: kernel_stack: <stack trace> => cpu_idle (ffffffff8100a38c) => start_secondary (ffffffff814ab828) trace-cmd-4132 [003] 64003.696632: function: schedule trace-cmd-4132 [003] 64003.696637: kernel_stack: <stack trace> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Convert macro opt args into enumsSteven Rostedt2011-03-11
| | | | Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Fold long option parsingLai Jiangshan2011-03-11
| | | | | | Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> LKML-Reference: <4D78A0A9.40000@cn.fujitsu.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Add -T option to perform stacktraces on each eventSteven Rostedt2011-03-11
| | | | | | | | | | | | | | | Add -T option to perform a stacktrace on each event: kworker/3:2-2393 [003] 58549.282806: sched_switch: kworker/3:2:2393 [120] S ==> trace-cmd:2603 [120] kworker/3:2-2393 [003] 58549.282808: kernel_stack: <stack trace> => schedule (ffffffff814b260e) => worker_thread (ffffffff8106edd7) => kthread (ffffffff81072b43) => kernel_thread_helper (ffffffff814bcbe4) <idle>-0 [000] 58549.282866: sched_switch: swapper:0 [120] R ==> trace-cmd:2600 [120] <idle>-0 [000] 58549.282869: kernel_stack: <stack trace> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Allow setting CC and AR, or CROSS_COMPILE from command lineDavid Sharp2011-03-10
| | | | | | | | | | | Makefiles suck: Use some makefile magic to get the best of all worlds for CC and AR: Sane defaults and the ability to override CC, and AR, or use CROSS_COMPILE to set a prefix. Signed-off-by: David Sharp <dhsharp@google.com> LKML-Reference: <1299791491-1805-1-git-send-email-dhsharp@google.com> Cc: Darren Hart <dvhart@linux.intel.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Correct cpu list parsingLai Jiangshan2011-03-10
| | | | | | | | | | | | | | | | | | | | Old parsing code is incorrect, when we execute ./trace-cmd report --cpu 1,2-4,6,9-13 the result of filter_cpus will be: 1,2,1,2,3,4,6,9,5,6,7,8,9,10,11,12,13 The correct filter_cpus should be: 1,2,3,4,6,9,10,11,12,13, I could change about 10 existing lines to make it correct, but the result would still be poorly readable, so I ported the neat and nicely readable code: linux_kernel_src/lib/bitmap.c:bitmap_parselist() to here and replaced the origin code. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> LKML-Reference: <4D78A0A6.2060009@cn.fujitsu.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Remove single use cpu code (use multi for all)Lai Jiangshan2011-03-10
| | | | | | | | | Remove the single use "filter_cpu" code and have one or multiple filtered cpus use the same code. Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com> LKML-Reference: <4D78A0A4.1040001@cn.fujitsu.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* parse-events: Support additional operators: '!', '~', and '!='David Sharp2011-03-09
| | | | | | | | | Add support for the unary operators '!' and '~', and support '!=' so that it is differentiated from '!'. Signed-off-by: David Sharp <dhsharp@google.com> LKML-Reference: <1299715137-22768-2-git-send-email-dhsharp@google.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* parse-events: Add support for printing short fieldsDavid Sharp2011-03-09
| | | | | | | | | Handle "%hd" etc. in pretty_print() Signed-off-by: David Sharp <dhsharp@google.com> LKML-Reference: <1299715137-22768-1-git-send-email-dhsharp@google.com> Google-Bug-Id: 3501052 Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* kernelshark: Have ^C on the command line kill the appSteven Rostedt2011-03-01
| | | | | | | | | The python plugin overrides the default kill of SIGINT. Have kernelshark add its own handler to kill it as well. TODO: Perhaps have a pop up to ask for sure? Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Remove duplicate search for ftrace event in blktrace hackSteven Rostedt2011-02-25
| | | | | | | | | | | | The first try of the blktrace hack, I searched for a ftrace event. But then I found that TRACE_BLK is not the same on all kernels so I did a progressive search for a ftrace event to find where TRACE_BLK may be. The linear search was left over from the first attempt and no longer needed. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Add hack to report out blktraceSteven Rostedt2011-02-25
| | | | | | | | | | | | The blktrace never exported the ftrace events via the /debug/tracing/events directory. Not to mention, that the blktrace is much more complex data to read out. Add a hack into the trace-input that creates a event format that parse-events can read for the blktrace file, and also create a plugin to parse the complex data. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Have writing options to file be endian awareSteven Rostedt2011-02-25
| | | | | | | | It is possible that the options written to file is not the same endian as the host endian. Have the options size and id converted first. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Fix freeing on error in tracecmd_attach_cpu_data_fd()Steven Rostedt2011-02-25
| | | | | | | tracecmd_attach_cpu_data_fd() had a few error paths that did not properly free memory. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Limit amount of events to send over the networkSteven Rostedt2011-02-25
| | | | | | | Limit the events sent out over the network the same way they are limited when being recorded locally to a file. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Add tracecmd_create_init_fd_glob()Steven Rostedt2011-02-25
| | | | | | | Add a way to create a trace.dat file from a file descriptor and limit the events that are used. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: extract command needs to record all eventsSteven Rostedt2011-02-25
| | | | | | | | | | When changing the code to limit the events recorded into the trace file to the events chosen by the user, it broke the extract code. The extract does not know what events are needed so it must record all events. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Record date first on record, last on extractSteven Rostedt2011-02-25
| | | | | | | | | | | | | | | Since record resets everything at the beginning, and may even keep the trace going when done (-k option), it is best to record the date before recording the trace. This way we can record the date just after resetting the buffers before starting the trace. Extract will record all the current data, so it is best to record the date when the extract is done, and then it is safe to reset the buffers to capture the date. We also ignore date with the start option. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Add trace-cmd record --date optionSteven Rostedt2011-02-24
| | | | | | | | | | | Adding the --date option to trace-cmd record (and extract) will cause trace-cmd to write a timestamp into the ftrace buffer and use it to mapped the ftrace buffer's timestamps with gettimeofday. This will allow reading of the trace.dat file to be able to convert the timestamps into actual wall time. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Close memory leaks in trace-output.cSteven Rostedt2011-02-24
| | | | | | Thanks again to valgrind. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Add tracecmd_create_init_file_glob()Steven Rostedt2011-02-24
| | | | | | | Add a version of tracecmd_create_init_file() that allows the caller to specify what events to save. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Add trace.dat option TRACECMD_OPTION_DATESteven Rostedt2011-02-24
| | | | | | | | | | Add the DATE option to trace.dat to store an ASCII hex number starting with '0x' that contains the difference between the timestamps stored in the trace.dat file and the gettimeofday value. The DATE options units is microseconds. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Fix writing options data to fileSteven Rostedt2011-02-24
| | | | | | | The address of the options data is written to the file instead of the actual data. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Add tracecmd_read_page_record()Steven Rostedt2011-02-24
| | | | | | | | Add new function tracecmd_read_page_record() that can parse records from a page without the need of a tracecmd_input handle. It requires a pevent, a page read from the ring buffer, the size of the page and the last_record. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Have translation functions not depend on tracecmd_inputSteven Rostedt2011-02-23
| | | | | | | Change the tracecmd input translation functions to only depend on the pevent instead of the tracecmd_input handle. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Fix while() indentation in trace-record.cSteven Rostedt2011-02-23
| | | | | | | Fix the while() loop indentation, which was only kept to help git see that the code was more of a move than any changes. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Move trace-cmd record code out of trace-cmd.cSteven Rostedt2011-02-23
| | | | | | Move the trace-cmd record code into its own file trace-record.c Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Rename trace-record.c to trace-recorder.cSteven Rostedt2011-02-23
| | | | | | | | | With the planning of creating a separate file for tracecmd record, the trace_recorder code should go into trace-recorder.c instead of trace-record.c. This way the tracecmd record code can have a file with the same name. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Allow multiple data files to be readSteven Rostedt2011-02-22
| | | | | | | | Let the user add -i multiple times to read multiple trace.dat files that will be merged sorted according to the timestamp that is given. Suggested-by: Thomas Taranowski <tom@baringforge.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* parse-events: Free token in processing function handlerSteven Rostedt2011-02-22
| | | | | | | | | A missing free_token() in the function handler of pevent was causing a memory leak. Thanks again to valgrind. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* parse-events: Only allocate comm when adding a new mappingSteven Rostedt2011-02-22
| | | | | | | | | Do not go through the malloc of a comm when checking to add a new one to the pid mapping until it is known that a new one is going to be added. Moving the alloction further down saves a little time. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* parse-events: Free comm in add_new_comm()Steven Rostedt2011-02-22
| | | | | | | | | | When the comm sent into add_new_comm() is not a match, we exit. What should be done is to free the comm first otherwise we leak memory. Found once again by valgrind. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* parse-events: Whitespace fixSteven Rostedt2011-02-22
| | | | Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Remove redundant type field in list_for_each_entry()Steven Rostedt2011-02-22
| | | | | | | The type field in list_for_each_entry() is the same as the type that the ptr field points to. Use typeof(*ptr) instead. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-graph: Change cursor when updating graphSteven Rostedt2011-02-21
| | | | | | When updating the graph, change the cursor to the wait cursor. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* kernelshark: Push and pop cursorSteven Rostedt2011-02-21
| | | | | | | | | Use a stack to save off the parent window cursor and restoring it via the stack. This way multiple trace_set_cursor() will work with multiple trace_put_cursor(). We can save and restore the cursor of the main window at various levels. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-graph: Free last_record in cpu plot on restartSteven Rostedt2011-02-21
| | | | | | | The restart of drawing a cpu plot may still have a left over record. Free it before initializing a new drawing. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-graph: Free last_record in cpu plotSteven Rostedt2011-02-21
| | | | | | When destroying a CPU plot, make sure the last record is freed. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-cmd: Add debugging to find where leaked records areSteven Rostedt2011-02-21
| | | | | | | | Add DEBUG_RECORD macro that records the address of the caller to get a record, and if there are leaked records existing, print out where the extra records were allocated from. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* kerenlshark: Only set cursor or freeze window if parent existsSteven Rostedt2011-02-21
| | | | | | If parent_window is NULL have both set_cursor and freeze do nothing. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* kernelshark: Place gdk lock around window showSteven Rostedt2011-02-21
| | | | | | | In order to be able to update the cursor while drawing windows, the first drawing of the main window requires the gdk_threads_enter(). Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-graph: Only unref plot->gc if it existsSteven Rostedt2011-02-21
| | | | | | | In case the loop breaks out early, we may unref plot->gc even though it does not exist. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-view: Do less gtk_main_iteration_do() in searchSteven Rostedt2011-02-21
| | | | | | | To speed up the search function slightly, only call the gtk_main_iteration_do() every 32 iterations. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
* trace-view: Search next pages if current page fails searchSteven Rostedt2011-02-21
| | | | | | | | If the list search fails to find the search text on the first page, ask to search on the other pages and continue the search if yes is selected. Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
generated by cgit v1.2.2 (git 2.25.0) at 2025-04-06 12:49:26 -0400