- 23 May, 2008 40 commits
-
-
Mathieu Desnoyers authored
Denys Vlasenko <vda.linux@googlemail.com> : > Not in this patch, but I noticed: > > #define __trace_mark(name, call_private, format, args...) \ > do { \ > static const char __mstrtab_##name[] \ > __attribute__((section("__markers_strings"))) \ > = #name "\0" format; \ > static struct marker __mark_##name \ > __attribute__((section("__markers"), aligned(8))) = \ > { __mstrtab_##name, &__mstrtab_##name[sizeof(#name)], \ > 0, 0, marker_probe_cb, \ > { __mark_empty_function, NULL}, NULL }; \ > __mark_check_format(format, ## args); \ > if (unlikely(__mark_##name.state)) { \ > (*__mark_##name.call) \ > (&__mark_##name, call_private, \ > format, ## args); \ > } \ > } while (0) > > In this call: > > (*__mark_##name.call) \ > (&__mark_##name, call_private, \ > format, ## args); \ > > you make gcc allocate duplicate format string. You can use > &__mstrtab_##name[sizeof(#name)] instead since it holds the same string, > or drop ", format," above and "const char *fmt" from here: > > void (*call)(const struct marker *mdata, /* Probe wrapper */ > void *call_private, const char *fmt, ...); > > since mdata->format is the same and all callees which need it can take it there. Very good point. I actually thought about dropping it, since it would remove an unnecessary argument from the stack. And actually, since I now have the marker_probe_cb sitting between the marker site and the callbacks, there is no API change required. Thanks :) Mathieu Signed-off-by:Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> CC: Denys Vlasenko <vda.linux@googlemail.com> Signed-off-by:
Ingo Molnar <mingo@elte.hu> Signed-off-by:
Thomas Gleixner <tglx@linutronix.de>
-
Steven Rostedt authored
Currently there is no protection from the root user to use up all of memory for trace buffers. If the root user allocates too many entries, the OOM killer might start kill off all tasks. This patch adds an algorith to check the following condition: pages_requested > (freeable_memory + current_trace_buffer_pages) / 4 If the above is met then the allocation fails. The above prevents more than 1/4th of freeable memory from being used by trace buffers. To determine the freeable_memory, I made determine_dirtyable_memory in mm/page-writeback.c global. Special thanks goes to Peter Zijlstra for suggesting the above calculation. Signed-off-by:
Steven Rostedt <srostedt@redhat.com> Signed-off-by:
-
Pekka Paalanen authored
Refactor code from tracing_read_pipe() and create trace_seq_to_user(). Moved trace_seq_reset() call before iter->trace->read() call so that when all leftover data is returned, trace_seq is reset automatically. Signed-off-by:
Pekka Paalanen <pq@iki.fi> Signed-off-by:
-
Steven Rostedt authored
Due to debug hooks in the kernel that can change the way smp_processor_id works, use raw_smp_processor_id in mcount called functions (namely ftrace_record_ip). Currently we annotate most debug functions from calling mcount, but we should not rely on that to prevent kernel lockups. This patch uses the raw_smp_processor_id to prevent a recusive crash that can happen if a debug hook in smp_processor_id calls mcount. Signed-off-by:
-
Ingo Molnar authored
Signed-off-by:
-
Ingo Molnar authored
Signed-off-by:
-
Steven Rostedt authored
In resetting the iterator in read_pipe, the reset of pos was postitioned in the wrong location with respect to the memset operation. The current code sets pos, incorrectly, to zero. Signed-off-by:
-
Pekka Paalanen authored
text_poke is sleepable. The original fix by Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>. Signed-off-by:
-
Steven Rostedt authored
This patch adds a method for open_pipe and open_read to the pluggins so that they can add a header to the trace pipe call. Signed-off-by:
-
Steven Rostedt authored
This patch sets up the infrastructure to record overruns of the tracing buffer. Signed-off-by:
-
Pekka Paalanen authored
Without CONFIG_DYNAMIC_FTRACE, mark_rodata_ro() would mark a wrong number of pages as no-execute. The bug was introduced in the patch "ftrace: dont write protect kernel text". The symptom was machine reboot after a CPU hotplug. Signed-off-by:
-
Steven Rostedt authored
In cleaning up of the sched_switch code, the function trace recording of task comms was removed. This patch adds back the recording of comms for function trace. The output of ftrace now has the task comm instead of <...>. Signed-off-by:
-
Steven Rostedt authored
This is first installment of adding documentation to the ftrace. Expect many more patches of this kind in the near future. Signed-off-by:
-
Steven Rostedt authored
Currently ftrace allocates a trace buffer for every possible CPU. Work is being done to change it to only online CPUs and add hooks to hotplug CPUS. This patch lays out the infrastructure for such a change. Signed-off-by:
-
Steven Rostedt authored
Andrew Morton suggested using strict_strtoul over simple_strtoul. This patch replaces them in ftrace. Signed-off-by:
-
Steven Rostedt authored
Andrew Morton mentioned some clean ups that should be done to ftrace. This patch does some of the simple clean ups. Signed-off-by:
-
Ingo Molnar authored
no code changed. Signed-off-by:
-
Thomas Gleixner authored
simplify hex to ascii conversion. Signed-off-by:
-
Steven Rostedt authored
With the introduction of ftrace, it is possible to recurse into the lockdep functions via the mcount call. To prevent possible lockups, updating the lockdep_recursion counter on grabbing the internal lockdep_lock should prevent deadlocks. Signed-off-by:
-
Steven Rostedt authored
This patch adds /debug/tracing/trace_entries that allows users to see as well as modify the number of trace entries the buffers hold. The number of entries only increments in ENTRIES_PER_PAGE which is calculated by the size of an entry with the number of entries that can fit in a page. The user does not need to use an exact size, but the entries will be rounded to one of the increments. Trying to set the entries to 0 will return with -EINVAL. To avoid race conditions, the modification of the buffer size can only be done when tracing is completely disabled (current_tracer == none). A info message will be printed if a user tries to modify the buffer size when not set to none. Signed-off-by:
-
Steven Rostedt authored
The wakeup selftest used an internal API for setting the test task priority. This patch fixes it to use the proper API for performing such a task. Thanks goes to Randy Dunlap for pointing out this build failure. Signed-off-by:
-
Steven Rostedt authored
This patch implements "NONBLOCK" for trace_pipe. If the trace_pipe is opened with O_NONBLOCK, then the trace_pipe read will not block when buffer is empty. Signed-off-by:
-
Steven Rostedt authored
Break out of while loop with EOF when the current_trace changes. Signed-off-by:
-
Steven Rostedt authored
The trace iterator is reset in the read. We still need to restore the tracer that the trace_pipe was opened with. Signed-off-by:
-
Steven Rostedt authored
We expect things like "cat" to block on reads to trace_pipe. That's what trace_pipe is for. Signed-off-by:
-
Ankita Garg authored
The conversion of task states to a character in the sched_switch tracer (part of latency tracer infrastructure), seems to be incorrect. We currently do it by indexing into the state_to_char array using the state value. The state values do not map directly into the array index and are thus incorrect. The following patch addresses this issue. This is also what is being done even in the show_task() routine in kernel/sched.c The patch has been compile and run tested. Signed-off-by:
Ankita Garg <ankita@in.ibm.com> Signed-off-by:
-
Thomas Gleixner authored
picked from the mmiotracer patches to distangle the patch queues. Signed-off-by: -
Pekka Paalanen authored
Provides kernel modules a way to register custom page fault handlers. On every page fault this will call a list of registered functions. The functions may handle the fault and force do_page_fault() to return immediately. This functionality is similar to the now removed page fault notifiers. Custom page fault handlers are used by debugging and reverse engineering tools. Mmiotrace is one such tool and a patch to add it into the tree will follow. The custom page fault handlers are called earlier in do_page_fault() than the page fault notifiers were. Signed-off-by:
-
Steven Rostedt authored
Dynamic ftrace cant work when the kernel has its text write protected. This patch keeps the kernel from being write protected when dynamic ftrace is in place. Signed-off-by:
-
Steven Rostedt authored
The fault label to jump to on fault of updating the code was misplaced preventing the fault from being recorded. Signed-off-by:
-
Steven Rostedt authored
There is a slight race condition in the selftest where the max update of the wakeup and irqs/preemption off tests can be doing a max update as the buffers are being tested. If this happens the system can crash with a GPF. This patch adds the max update spinlock around the checking of the buffers to prevent such a race. Signed-off-by:
-
Steven Rostedt authored
If the trace output changes on reading the trace files, there is a chance that the start function will return NULL. If the start function of a sequence returns NULL the stop equivalent is not called. In this case, all locks that are taken must be released even if they are released in the stop function. This patch fixes a case that a mutex was not released on return of NULL in the start sequence function. Signed-off-by:
-
Steven Rostedt authored
When dynamic ftrace fails and sets itself disabled, the ftraced daemon will go back to sleep everytime it wakes up. The setting of the ftraced state to UNINTERRUPTIBLE is skipped in this process, and the daemon takes up 100% of the CPU. This patch makes sure the ftraced daemon sets itself to UNINTERRUPTIBLE in that loop. Signed-off-by:
-
Ingo Molnar authored
Signed-off-by:
-
Steven Rostedt authored
This patch removes the "notrace" annotation from lockdep and adds the debugging files in the kernel director to those that should not be compiled with "-pg" mcount tracing. Signed-off-by:
-
Steven Rostedt authored
The debug functions in spin_lock debugging pollute the output of the function tracer. This patch adds the debug files in the lib director to those that should not be compiled with mcount tracing. Signed-off-by:
-
Steven Rostedt authored
Lock debugging enabled cause huge performance problems for tracing. Having the lock verification happening for every function that is called because mcount calls spin_lock can cripple the system. This patch converts the spin_locks used by ftrace into raw_spin_locks. Signed-off-by:
-
Steven Rostedt authored
This patch changes the use of __get_cpu_var to explicitly calling raw_smp_processor_id and using the per_cpu() macro. On some debug configurations, the use of __get_cpu_var may cause ftrace to trigger and this can cause problems with the irqsoff tracing. Signed-off-by:
-
Ingo Molnar authored
Signed-off-by:
-
Steven Rostedt authored
With the adding of the configuration changes in the Makefile to prevent tracing of functions in the ftrace code, all tracing of all the ftrace code has been removed. Unfortunately, one of the selftests, relied on a function to be traced. With the new change, the function was no longer traced and the test failed. This patch separates out the test function into its own file so that we can add the "-pg" flag to the compilation of that function and the adding of the mcount call to that function. Signed-off-by:
-
