- 25 Sep, 2018 33 commits
-
-
Suzuki K Poulose authored
Add support for reporting errors back from the SMP cross function call for enabling ETM. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
Suzuki K Poulose <suzuki.poulose@arm.com> Signed-off-by:
Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Suzuki K Poulose authored
Add support for handling errors in enabling the component. The ETM is enabled via cross call to owner CPU. Make necessary changes to report the error back from the cross call. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
Prepare to handle errors in enabling the hardware and report it back to the core driver. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
Make sure we honor the errors in CATU device and abort the operation. While at it, delay setting the etr_buf for the session until we are sure that we are indeed enabling the ETR. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
Refactor the tmc-etr enable operation to make it easier to handle errors in enabling the hardware. We need to make sure that the buffer is compatible with the ETR. This patch re-arranges to make the error handling easier, by deferring the hardware enablement until all the errors are checked. This also avoids turning the CATU on/off during a sysfs read session. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
coresight_enable_path() enables the components in a trace path from a given source to a sink, excluding the source. The operation is performed in the reverse order; the sink first and then backwards in the list. However, if we encounter an error in enabling any of the component, we simply disable all the components in the given path irrespective of whether we enabled some of the components in the enable iteration. This could interfere with another trace session if one of the link devices is turned off (e.g, TMC-ETF). So, we need to make sure that we only disable those components which were actually enabled from the iteration. This patch achieves the same by refactoring the coresight_disable_path to accept a "node" to start from in the forward order, which can then be used from the error path of coresight_enable_path(). With this change, we don't issue a disable call back for a component which didn't get enabled. This change of behavior triggers a bug in coresight_enable_link(), where we leave the refcount on the device and will prevent the device from being enabled forever. So, we also drop the refcount in the coresight_enable_link() if the operation failed. Also, with the refactoring, we always start after the first node (which is the "SOURCE" device) for disabling the entire path. This implies, we must not find a "SOURCE" in the middle of the path. Hence, added a WARN_ON() to make sure the paths we get are sane, rather than simply ignoring them. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Leo Yan authored
>From the comment in the code, it claims the requirement for byte-address alignment for RRP register: 'for 32-bit, 64-bit and 128-bit wide trace memory, the four LSBs must be 0s. For 256-bit wide trace memory, the five LSBs must be 0s'. This isn't consistent with the program, the program sets five LSBs as zeros for 32/64/128-bit wide trace memory and set six LSBs zeros for 256-bit wide trace memory. After checking with the CoreSight Trace Memory Controller technical reference manual (ARM DDI 0461B, section 3.3.4 RAM Read Pointer Register), it proves the comment is right and the program does wrong setting. This patch fixes byte-address alignment for RRP by following correct definition in the technical reference manual. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Cc: Mike Leach <mike.leach@linaro.org> Signed-off-by:
Leo Yan <leo.yan@linaro.org> Signed-off-by:
-
Leo Yan authored
In ETB dump function tmc_etb_dump_hw() it has nested loops. The second level loop is to iterate index in the range [0 .. drvdata->memwidth); but the index isn't really used in the code, thus the second level loop is useless. This patch is to remove the second level loop; the refactor also reduces indentation and we can use 'break' to replace 'goto' tag. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Tomasz Nowicki authored
For non-VHE systems host kernel runs at EL1 and jumps to EL2 whenever hypervisor code should be executed. In this case ETM4x driver must restrict configuration to EL1 when it setups kernel tracing. However, there is no separate hypervisor privilege level when VHE is enabled, the host kernel runs at EL2. This patch fixes configuration of TRCACATRn register for VHE systems so that ETM_EXLEVEL_NS_HYP bit is used instead of ETM_EXLEVEL_NS_OS to on/off kernel tracing. At the same time, it moves common code to new helper. Signed-off-by:
Tomasz Nowicki <tnowicki@caviumnetworks.com> Signed-off-by:
-
Mathieu Poirier authored
Up until now the relative simplicity of enabling the ETB made it possible to accommodate processing for both sysFS and perf methods. But work on claimtags and CPU-wide trace scenarios is adding some complexity, making the current code messy and hard to maintain. As such follow what has been done for ETF and ETR components and split function etb_enable() so that processing for both API can be done cleanly. Signed-off-by:
-
Mathieu Poirier authored
This patch moves the etb_drvdata::mode from a locat_t to a simple u32, as it is for the ETF and ETR drivers. This streamlines the code and adds commonality with the other drivers when dealing with similar operations. Signed-off-by:
-
Suzuki K Poulose authored
Add support for using TMC-ETR as backend for ETM perf tracing. We use software double buffering at the moment. i.e, the TMC-ETR uses a separate buffer than the perf ring buffer. The data is copied to the perf ring buffer once a session completes. The TMC-ETR would try to match the larger of perf ring buffer or the ETR buffer size configured via sysfs, scaling down to a minimum limit of 1MB. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
In coresight perf mode, we need to prepare the sink before starting a session, which is done via set_buffer call back. We then proceed to enable the tracing. If we fail to start the session successfully, we leave the sink configuration unchanged. In order to make the operation atomic and to avoid yet another call back to clear the buffer, we get rid of the "set_buffer" call back and pass the buffer details via enable() call back to the sink. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
We can always find the sink configuration for a given perf_output_handle. Add a helper to retrieve the sink configuration for a given perf_output_handle. This will be used to get rid of the set_buffer() call back. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
Right now we issue an update_buffer() and reset_buffer() call backs in succession when we stop tracing an event. The update_buffer is supposed to check the status of the buffer and make sure the ring buffer is updated with the trace data. And we store information about the size of the data collected only to be consumed by the reset_buffer callback which always follows the update_buffer. This was originally designed for handling future IPs which could trigger a buffer overflow interrupt. This patch gets rid of the reset_buffer callback altogether and performs the actions in update_buffer, making it return the size collected. We can always add the support for handling the overflow interrupt case later. This removes some not-so pretty hack (storing the new head in the size field for snapshot mode) and cleans it up a little bit. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
Convert component enable/disable messages from dev_info to dev_dbg. When used with perf, the components in the paths are enabled/disabled during each schedule of the run, which can flood the dmesg with these messages. Moreover, they are only useful for debug purposes. So, convert such messages to dev_dbg() which can be turned on as needed. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
Since the ETR now uses mode specific buffers, we can reliably provide the trace data captured in sysfs mode, even when the ETR is operating in PERF mode. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
Since the ETR could be driven either by SYSFS or by perf, it becomes complicated how we deal with the buffers used for each of these modes. The ETR driver cannot simply free the current attached buffer without knowing the provider (i.e, sysfs vs perf). To solve this issue, we provide: 1) the driver-mode specific etr buffer to be retained in the drvdata 2) the etr_buf for a session should be passed on when enabling the hardware, which will be stored in drvdata->etr_buf. This will be replaced (not free'd) as soon as the hardware is disabled, after necessary sync operation. The advantages of this are : 1) The common code path doesn't need to worry about how to dispose an existing buffer, if it is about to start a new session with a different buffer, possibly in a different mode. 2) The driver mode can control its buffers and can get access to the saved session even when the hardware is operating in a different mode. (e.g, we can still access a trace buffer from a sysfs mode even if the etr is now used in perf mode, without disrupting the current session.) Towards this, we introduce a sysfs specific data which will hold the etr_buf used for sysfs mode of operation, controlled solely by the sysfs mode handling code. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
We enable the trace path, before activating the source. If we fail to enable the source, we must disable the path to make sure it is available for another session. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
At the moment, if there is no CPU specified for a given event, we use cpu_online_mask and try to build path for each of the CPUs in the mask. This could prevent any CPU that is turned online later to be used for the tracing. This patch changes to use the cpu_present_mask and tries to build path for as much CPUs as possible ignoring the failures in building path for some of the CPUs. If ever we try to trace on those CPUs, we fail the operation. Based on a patch from Mathieu Poirier. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
We hold the read lock on CPU hotplug to simply copy the online mask, which is not really needed. And this can cause a lockdep warning, like : [ 54.632093] ====================================================== [ 54.638207] WARNING: possible circular locking dependency detected [ 54.644322] 4.18.0-rc3-00042-g2d39e6356bb7-dirty #309 Not tainted [ 54.650350] ------------------------------------------------------ [ 54.656464] perf/2862 is trying to acquire lock: [ 54.661031] 000000007e21d170 (&event->mmap_mutex){+.+.}, at: perf_event_set_output+0x98/0x138 [ 54.669486] [ 54.669486] but task is already holding lock: [ 54.675256] 000000001080eb1b (&cpuctx_mutex){+.+.}, at: perf_event_ctx_lock_nested+0xf8/0x1f0 [ 54.683704] [ 54.683704] which lock already depends on the new lock. [ 54.683704] [ 54.691797] [ 54.691797] the existing dependency chain (in reverse order) is: [ 54.699201] [ 54.699201] -> #3 (&cpuctx_mutex){+.+.}: [ 54.704556] __mutex_lock+0x70/0x808 [ 54.708608] mutex_lock_nested+0x1c/0x28 [ 54.713005] perf_event_init_cpu+0x8c/0xd8 [ 54.717574] perf_event_init+0x194/0x1d4 [ 54.721971] start_kernel+0x2b8/0x42c [ 54.726107] [ 54.726107] -> #2 (pmus_lock){+.+.}: [ 54.731114] __mutex_lock+0x70/0x808 [ 54.735165] mutex_lock_nested+0x1c/0x28 [ 54.739560] perf_event_init_cpu+0x30/0xd8 [ 54.744129] cpuhp_invoke_callback+0x84/0x248 [ 54.748954] _cpu_up+0xe8/0x1c8 [ 54.752576] do_cpu_up+0xa8/0xc8 [ 54.756283] cpu_up+0x10/0x18 [ 54.759731] smp_init+0xa0/0x114 [ 54.763438] kernel_init_freeable+0x120/0x288 [ 54.768264] kernel_init+0x10/0x108 [ 54.772230] ret_from_fork+0x10/0x18 [ 54.776279] [ 54.776279] -> #1 (cpu_hotplug_lock.rw_sem){++++}: [ 54.782492] cpus_read_lock+0x34/0xb0 [ 54.786631] etm_setup_aux+0x5c/0x308 [ 54.790769] rb_alloc_aux+0x1ec/0x300 [ 54.794906] perf_mmap+0x284/0x610 [ 54.798787] mmap_region+0x388/0x570 [ 54.802838] do_mmap+0x344/0x4f8 [ 54.806544] vm_mmap_pgoff+0xe4/0x110 [ 54.810682] ksys_mmap_pgoff+0xa8/0x240 [ 54.814992] sys_mmap+0x18/0x28 [ 54.818613] el0_svc_naked+0x30/0x34 [ 54.822661] [ 54.822661] -> #0 (&event->mmap_mutex){+.+.}: [ 54.828445] lock_acquire+0x48/0x68 [ 54.832409] __mutex_lock+0x70/0x808 [ 54.836459] mutex_lock_nested+0x1c/0x28 [ 54.840855] perf_event_set_output+0x98/0x138 [ 54.845680] _perf_ioctl+0x2a0/0x6a0 [ 54.849731] perf_ioctl+0x3c/0x68 [ 54.853526] do_vfs_ioctl+0xb8/0xa20 [ 54.857577] ksys_ioctl+0x80/0xb8 [ 54.861370] sys_ioctl+0xc/0x18 [ 54.864990] el0_svc_naked+0x30/0x34 [ 54.869039] [ 54.869039] other info that might help us debug this: [ 54.869039] [ 54.876960] Chain exists of: [ 54.876960] &event->mmap_mutex --> pmus_lock --> &cpuctx_mutex [ 54.876960] [ 54.887217] Possible unsafe locking scenario: [ 54.887217] [ 54.893073] CPU0 CPU1 [ 54.897552] ---- ---- [ 54.902030] lock(&cpuctx_mutex); [ 54.905396] lock(pmus_lock); [ 54.910911] lock(&cpuctx_mutex); [ 54.916770] lock(&event->mmap_mutex); [ 54.920566] [ 54.920566] *** DEADLOCK *** [ 54.920566] [ 54.926424] 1 lock held by perf/2862: [ 54.930042] #0: 000000001080eb1b (&cpuctx_mutex){+.+.}, at: perf_event_ctx_lock_nested+0xf8/0x1f0 Since we have per-cpu array for the paths, we simply don't care about the number of online CPUs. This patch gets rid of the {get/put}_online_cpus(). Reported-by: -
Suzuki K Poulose authored
We create a coresight trace path for each online CPU when we start the event. We rely on the number of online CPUs and then go on to allocate an array matching the "number of online CPUs" for holding the path and then uses normal CPU id as the index to the array. This is problematic as we could have some offline CPUs causing us to access beyond the actual array size (e.g, on a dual SMP system, if CPU0 is offline, CPU1 could be really accessing beyond the array). The solution is to switch to per-cpu array for holding the path. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
If the ETB is already enabled in sysfs mode, the ETB reports success even if a perf mode is requested. Fix this by checking the requested mode. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
The coresight components could be operated either in sysfs mode or in perf mode. For some of the components, the mode of operation doesn't matter as they simply relay the data to the next component in the trace path. But for sinks, they need to be able to provide the trace data back to the user. Thus we need to make sure that "mode" is handled appropriately. e.g, the sysfs mode could have multiple sources driving the trace data, while perf mode doesn't allow sharing the sink. The coresight_enable_sink() however doesn't really allow this check to trigger as it skips the "enable_sink" callback if the component is already enabled, irrespective of the mode. This could cause mixing of data from different modes or even same mode (in perf), if the sources are different. Also, if we fail to enable the sink while enabling a path (where sink is the first component enabled), we could end up in disabling the components in the "entire" path which were not enabled in this trial, causing disruptions in the existing trace paths. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
zhong jiang authored
Use ERR_CAT inlined function to replace the ERR_PTR(PTR_ERR). It make the code more concise. Signed-off-by:
zhong jiang <zhongjiang@huawei.com> Signed-off-by:
-
Suzuki K Poulose authored
The coresight drivers relied on default bindings for graph in DT, while reusing the "reg" field of the "ports" to indicate the actual hardware port number for the connections. This can cause duplicate ports with same addresses, but different direction. However, with the rules getting stricter for the address mismatch with the label, it is no longer possible to use the port address field for the hardware port number. This patch introduces new DT binding rules for coresight components, based on the same generic DT graph bindings, but avoiding the address duplication. - All output ports must be specified under a child node with name "out-ports". - All input ports must be specified under a childe node with name "in-ports". - Port address should match the hardware port number. The support for legacy bindings is retained, with a warning. Cc: Sudeep Holla <sudeep.holla@arm.com> Cc: Rob Herring <robh@kernel.org> Signed-off-by:
Rob Herring <robh@kernel.org> Signed-off-by:
-
Suzuki K Poulose authored
The platform code parses the component connections and populates a platform-description of the output connections in arrays of fields (which is never freed). This is later copied in the coresight_register to a newly allocated area, represented by coresight_connection(s). This patch cleans up the code dealing with connections by making use of the "coresight_connection" structure right at the platform code and lets the generic driver simply re-use information provided by the platform. Thus making it reader friendly as well as avoiding the wastage of unused memory. Signed-off-by:
-
Suzuki K Poulose authored
Add a helper to check if the given endpoint is input. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
When parsing the remote endpoint of an output port, we do : rport = of_graph_get_remote_port(ep); rparent = of_graph_get_remote_port_parent(ep); and then parse the "remote_port" as if it was the remote endpoint, which is wrong. The code worked fine because we used endpoint number as the port number. Let us fix it and optimise a bit as: remote_ep = of_graph_get_remote_endpoint(ep); if (remote_ep) remote_parent = of_graph_get_port_parent(remote_ep); and then, parse the remote_ep for the port/endpoint details. Signed-off-by: -
Suzuki K Poulose authored
We don't drop the reference on the remote device while parsing the connection, held by bus_find_device(). Fix this by duplicating the device name and dropping the reference. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Cc: Kim Phillips <kim.phillips@arm.com> Signed-off-by:
-
Suzuki K Poulose authored
The coresight driver doesn't drop the references on the remote endpoint/port nodes. Add the missing of_node_put() calls. Reported-by:
-
Suzuki K Poulose authored
Refactor the of graph endpoint parsing code, to make the error handling easier. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Signed-off-by:
-
Suzuki K Poulose authored
commit 6403587a930c ("coresight: use put_device() instead of kfree()") fixes the double freeing of resources and ensures that the device refcount is dropped properly. Add a comment to explain this to help the readers and prevent people trying to "unfix" it again. While at it, rename the labels for better readability. Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Cc: Arvind Yadav <arvind.yadav.cs@gmail.com> Signed-off-by:
-
- 20 Sep, 2018 1 commit
-
-
Gustavo A. R. Silva authored
Replace "fallthru" with a proper "fall through" annotation. This fix is part of the ongoing efforts to enabling -Wimplicit-fallthrough Signed-off-by:
Gustavo A. R. Silva <gustavo@embeddedor.com> Acked-by:
Dimitri Sivanich <sivanich@hpe.com> Signed-off-by:
-
- 16 Sep, 2018 3 commits
-
-
Greg Kroah-Hartman authored
We want the bugfixes in here as well. Signed-off-by: -
Linus Torvalds authored
-
Greg Kroah-Hartman authored
The Code of Conflict is not achieving its implicit goal of fostering civility and the spirit of 'be excellent to each other'. Explicit guidelines have demonstrated success in other projects and other areas of the kernel. Here is a Code of Conduct statement for the wider kernel. It is based on the Contributor Covenant as described at www.contributor-covenant.org From this point forward, we should abide by these rules in order to help make the kernel community a welcoming environment to participate in. Signed-off-by:
Chris Mason <clm@fb.com> Signed-off-by:
Dan Williams <dan.j.williams@intel.com> Signed-off-by:
Jonathan Corbet <corbet@lwn.net> Signed-off-by:
Olof Johansson <olof@lxom.net> Signed-off-by:
Steven Rostedt (VMware) <rostedt@goodmis.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
- 15 Sep, 2018 3 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull x86 fixes from Ingol Molnar: "Misc fixes: - EFI crash fix - Xen PV fixes - do not allow PTI on 2-level 32-bit kernels for now - documentation fix" * 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: x86/APM: Fix build warning when PROC_FS is not enabled Revert "x86/mm/legacy: Populate the user page-table with user pgd's" x86/efi: Load fixmap GDT in efi_call_phys_epilog() before setting %cr3 x86/xen: Disable CPU0 hotplug for Xen PV x86/EISA: Don't probe EISA bus for Xen PV guests x86/doc: Fix Documentation/x86/earlyprintk.txt
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull scheduler fixes from Ingo Molnar: "Misc fixes: various scheduler metrics corner case fixes, a sched_features deadlock fix, and a topology fix for certain NUMA systems" * 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: sched/fair: Fix kernel-doc notation warning sched/fair: Fix load_balance redo for !imbalance sched/fair: Fix scale_rt_capacity() for SMT sched/fair: Fix vruntime_normalized() for remote non-migration wakeup sched/pelt: Fix update_blocked_averages() for RT and DL classes sched/topology: Set correct NUMA topology type sched/debug: Fix potential deadlock when writing to sched_features
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull perf fixes from Ingo Molnar: "Mostly tooling fixes, but also breakpoint and x86 PMU driver fixes" * 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (23 commits) perf tools: Fix maps__find_symbol_by_name() tools headers uapi: Update tools's copy of linux/if_link.h tools headers uapi: Update tools's copy of linux/vhost.h tools headers uapi: Update tools's copies of kvm headers tools headers uapi: Update tools's copy of drm/drm.h tools headers uapi: Update tools's copy of asm-generic/unistd.h tools headers uapi: Update tools's copy of linux/perf_event.h perf/core: Force USER_DS when recording user stack data perf/UAPI: Clearly mark __PERF_SAMPLE_CALLCHAIN_EARLY as internal use perf/x86/intel: Add support/quirk for the MISPREDICT bit on Knights Landing CPUs perf annotate: Fix parsing aarch64 branch instructions after objdump update perf probe powerpc: Ignore SyS symbols irrespective of endianness perf event-parse: Use fixed size string for comms perf util: Fix bad memory access in trace info. perf tools: Streamline bpf examples and headers installation perf evsel: Fix potential null pointer dereference in perf_evsel__new_idx() perf arm64: Fix include path for asm-generic/unistd.h perf/hw_breakpoint: Simplify breakpoint enable in perf_event_modify_breakpoint perf/hw_breakpoint: Enable breakpoint in modify_user_hw_breakpoint perf/hw_breakpoint: Remove superfluous bp->attr.disabled = 0 ...
-
