BMC/Intel-BMC/linux.git/arch/x86/kernel/cpu/perf_event_intel_pt.c, branch dev

Merge tag 'module_init-alternate_initcall-v4.1-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/paulg/linux

2015-07-02T17:36:29+00:00

Pull module_init replacement part two from Paul Gortmaker: "Replace module_init with appropriate alternate initcall in non modules. This series converts non-modular code that is using the module_init() call to hook itself into the system to instead use one of our alternate priority initcalls. Unlike the previous series that used device_initcall and hence was a runtime no-op, these commits change to one of the alternate initcalls, because (a) we have them and (b) it seems like the right thing to do. For example, it would seem logical to use arch_initcall for arch specific setup code and fs_initcall for filesystem setup code. This does mean however, that changes in the init ordering will be taking place, and so there is a small risk that some kind of implicit init ordering issue may lie uncovered. But I think it is still better to give these ones sensible priorities than to just assign them all to device_initcall in order to exactly preserve the old ordering. Thad said, we have already made similar changes in core kernel code in commit c96d6660dc65 ("kernel: audit/fix non-modular users of module_init in core code") without any regressions reported, so this type of change isn't without precedent. It has also got the same local testing and linux-next coverage as all the other pull requests that I'm sending for this merge window have got. Once again, there is an unused module_exit function removal that shows up as an outlier upon casual inspection of the diffstat" * tag 'module_init-alternate_initcall-v4.1-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/paulg/linux: x86: perf_event_intel_pt.c: use arch_initcall to hook in enabling x86: perf_event_intel_bts.c: use arch_initcall to hook in enabling mm/page_owner.c: use late_initcall to hook in enabling lib/list_sort: use late_initcall to hook in self tests arm: use subsys_initcall in non-modular pl320 IPC code powerpc: don't use module_init for non-modular core hugetlb code powerpc: use subsys_initcall for Freescale Local Bus x86: don't use module_init for non-modular core bootflag code netfilter: don't use module_init/exit in core IPV4 code fs/notify: don't use module_init for non-modular inotify_user code mm: replace module_init usages with subsys_initcall in nommu.c

Merge branch 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

2015-06-22T22:19:21+00:00

Pull perf updates from Ingo Molnar: "Kernel side changes mostly consist of work on x86 PMU drivers: - x86 Intel PT (hardware CPU tracer) improvements (Alexander Shishkin) - x86 Intel CQM (cache quality monitoring) improvements (Thomas Gleixner) - x86 Intel PEBSv3 support (Peter Zijlstra) - x86 Intel PEBS interrupt batching support for lower overhead sampling (Zheng Yan, Kan Liang) - x86 PMU scheduler fixes and improvements (Peter Zijlstra) There's too many tooling improvements to list them all - here are a few select highlights: 'perf bench': - Introduce new 'perf bench futex' benchmark: 'wake-parallel', to measure parallel waker threads generating contention for kernel locks (hb->lock). (Davidlohr Bueso) 'perf top', 'perf report': - Allow disabling/enabling events dynamicaly in 'perf top': a 'perf top' session can instantly become a 'perf report' one, i.e. going from dynamic analysis to a static one, returning to a dynamic one is possible, to toogle the modes, just press 'f' to 'freeze/unfreeze' the sampling. (Arnaldo Carvalho de Melo) - Make Ctrl-C stop processing on TUI, allowing interrupting the load of big perf.data files (Namhyung Kim) 'perf probe': (Masami Hiramatsu) - Support glob wildcards for function name - Support $params special probe argument: Collect all function arguments - Make --line checks validate C-style function name. - Add --no-inlines option to avoid searching inline functions - Greatly speed up 'perf probe --list' by caching debuginfo. - Improve --filter support for 'perf probe', allowing using its arguments on other commands, as --add, --del, etc. 'perf sched': - Add option in 'perf sched' to merge like comms to lat output (Josef Bacik) Plus tons of infrastructure work - in particular preparation for upcoming threaded perf report support, but also lots of other work - and fixes and other improvements. See (much) more details in the shortlog and in the git log" * 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (305 commits) perf tools: Configurable per thread proc map processing time out perf tools: Add time out to force stop proc map processing perf report: Fix sort__sym_cmp to also compare end of symbol perf hists browser: React to unassigned hotkey pressing perf top: Tell the user how to unfreeze events after pressing 'f' perf hists browser: Honour the help line provided by builtin-{top,report}.c perf hists browser: Do not exit when 'f' is pressed in 'report' mode perf top: Replace CTRL+z with 'f' as hotkey for enable/disable events perf annotate: Rename source_line_percent to source_line_samples perf annotate: Display total number of samples with --show-total-period perf tools: Ensure thread-stack is flushed perf top: Allow disabling/enabling events dynamicly perf evlist: Add toggle_enable() method perf trace: Fix race condition at the end of started workloads perf probe: Speed up perf probe --list by caching debuginfo perf probe: Show usage even if the last event is skipped perf tools: Move libtraceevent dynamic list to separated LDFLAGS variable perf tools: Fix a problem when opening old perf.data with different byte order perf tools: Ignore .config-detected in .gitignore perf probe: Fix to return error if no probe is added ...

x86: perf_event_intel_pt.c: use arch_initcall to hook in enabling

2015-06-16T18:12:35+00:00

This was using module_init, but the current Kconfig situation is as follows: In arch/x86/kernel/cpu/Makefile: obj-$(CONFIG_CPU_SUP_INTEL) += perf_event_intel_pt.o perf_event_intel_bts.o and in arch/x86/Kconfig.cpu: config CPU_SUP_INTEL default y bool "Support Intel processors" if PROCESSOR_SELECT So currently, the end user can not build this code into a module. If in the future, there is desire for this to be modular, then it can be changed to include and use module_init. But currently, in the non-modular case, a module_init becomes a device_initcall. But this really isn't a device, so we should choose a more appropriate initcall bucket to put it in. The obvious choice here seems to be arch_initcall, but that does make it earlier than it was currently through device_initcall. As long as perf_pmu_register() is functional, we should be OK. Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Ingo Molnar Cc: Arnaldo Carvalho de Melo Cc: Thomas Gleixner Cc: "H. Peter Anvin" Cc: x86@kernel.org Signed-off-by: Paul Gortmaker

perf/x86/intel/pt: Fix a refactoring bug

2015-06-04T14:07:51+00:00

Commit 066450be41 ("perf/x86/intel/pt: Clean up the control flow in pt_pmu_hw_init()") changed attribute initialization so that only the first attribute gets initialized using sysfs_attr_init(), which upsets lockdep. This patch fixes the glitch so that all allocated attributes are properly initialized thus fixing the lockdep warning reported by Tvrtko and Imre. Reported-by: Tvrtko Ursulin Reported-by: Imre Deak Signed-off-by: Alexander Shishkin Cc: Cc: Andrew Morton Cc: H. Peter Anvin Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Signed-off-by: Ingo Molnar

perf/x86/intel/pt: Remove redundant variable declaration

2015-05-27T07:17:48+00:00

There is a 'pt' variable in the outer scope of pt_event_stop() with the same type, we don't really need another one in the inner scope. This patch removes the redundant variable declaration. Signed-off-by: Alexander Shishkin Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: acme@infradead.org Cc: adrian.hunter@intel.com Cc: hpa@zytor.com Link: http://lkml.kernel.org/r/1432308626-18845-8-git-send-email-alexander.shishkin@linux.intel.com Signed-off-by: Ingo Molnar

perf/x86/intel/pt: Kill pt_is_running()

2015-05-27T07:17:48+00:00

Initially, we were trying to guard against scenarios where somebody attaches to the system with a hardware debugger while PT is enabled from software and pt_is_running() tries to make sure we handle this better, but the truth is, there is still a race window no matter what and people with hardware debuggers should really know what they are doing anyway. In other words, there is no point in keeping this one around, and it's one RDMSR instructions fewer in the fast path. The case when PT is enabled by the BIOS at boot time is handled in the driver initialization path and doesn't use pt_is_running(). This patch gets rid of it. Signed-off-by: Alexander Shishkin Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: acme@infradead.org Cc: adrian.hunter@intel.com Cc: hpa@zytor.com Link: http://lkml.kernel.org/r/1429622177-22843-6-git-send-email-alexander.shishkin@linux.intel.com Signed-off-by: Ingo Molnar

perf/x86/intel/pt: Document pt_buffer_reset_offsets()

2015-05-27T07:17:47+00:00

Currently, the description of pt_buffer_reset_offsets() lacks information about its calling constraints and ordering with regards to other buffer management functions. Add a clarification about when this function has to be called. Signed-off-by: Alexander Shishkin Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: acme@infradead.org Cc: adrian.hunter@intel.com Cc: hpa@zytor.com Link: http://lkml.kernel.org/r/1429622177-22843-5-git-send-email-alexander.shishkin@linux.intel.com Signed-off-by: Ingo Molnar

perf/x86/intel/pt: Document pt_buffer_reset_markers()

2015-05-27T07:17:47+00:00

The comments in the driver don't make it absolutely clear as to what exactly is the calling order and other possible constraints of buffer management functions. Document constraints and calling order for the buffer configuration functions. While at it, replace a redundant check in pt_buffer_reset_markers() with an explanation why it is not needed. Signed-off-by: Alexander Shishkin Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: acme@infradead.org Cc: adrian.hunter@intel.com Cc: hpa@zytor.com Link: http://lkml.kernel.org/r/1429622177-22843-4-git-send-email-alexander.shishkin@linux.intel.com Signed-off-by: Ingo Molnar

perf/x86/intel/pt: Kill an unused variable

2015-05-27T07:17:46+00:00

Currently, there's a set-but-not-used variable in setup_topa_index(); this patch gets rid of it. And while at it, fixes a style issue with brackets around a one-line block. Signed-off-by: Alexander Shishkin Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: acme@infradead.org Cc: adrian.hunter@intel.com Cc: hpa@zytor.com Link: http://lkml.kernel.org/r/1429622177-22843-2-git-send-email-alexander.shishkin@linux.intel.com Signed-off-by: Ingo Molnar

perf/x86/intel/pt: Untangle pt_buffer_reset_markers()

2015-05-27T07:16:20+00:00

Currently, pt_buffer_reset_markers() is a difficult to read knot of arithmetics with a redundant check for multiple-entry TOPA capability, a commented out wakeup marker placement and a logical error wrt to stop marker placement. The latter happens when write head is not page aligned and results in stop marker being placed one page earlier than it actually should. All these problems only affect PT implementations that support multiple-entry TOPA tables (read: proper scatter-gather). For single-entry TOPA implementations, there is no functional impact. This patch deals with all of the above. Tested on both single-entry and multiple-entry TOPA PT implementations. Signed-off-by: Alexander Shishkin Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: acme@infradead.org Cc: adrian.hunter@intel.com Cc: hpa@zytor.com Link: http://lkml.kernel.org/r/1432308626-18845-4-git-send-email-alexander.shishkin@linux.intel.com Signed-off-by: Ingo Molnar