Changes in 4.9.233 xfs: catch inode allocation state mismatch corruption xfs: validate cached inodes are free when allocated xfs: don't call xfs_da_shrink_inode with NULL bp net: phy: mdio-bcm-unimac: fix potential NULL dereference in unimac_mdio_probe() crypto: ccp - Release all allocated memory if sha type is invalid media: rc: prevent memory leak in cx23888_ir_probe ath9k_htc: release allocated buffer if timed out ath9k: release allocated buffer if timed out PCI/ASPM: Disable ASPM on ASMedia ASM1083/1085 PCIe-to-PCI bridge ARM: 8986/1: hw_breakpoint: Don't invoke overflow handler on uaccess watchpoints drm/amdgpu: Prevent kernel-infoleak in amdgpu_info_ioctl() drm: hold gem reference until object is no longer accessed f2fs: check memory boundary by insane namelen f2fs: check if file namelen exceeds max value 9p/trans_fd: abort p9_read_work if req status changed 9p/trans_fd: Fix concurrency del of req_list in p9_fd_cancelled/p9_read_work x86/build/lto: Fix truncated .bss with -fdata-sections x86, vmlinux.lds: Page-align end of ..page_aligned sections fbdev: Detect integer underflow at "struct fbcon_ops"->clear_margins. rds: Prevent kernel-infoleak in rds_notify_queue_get() xfs: fix missed wakeup on l_flush_wait uapi: includes linux/types.h before exporting files install several missing uapi headers net/x25: Fix x25_neigh refcnt leak when x25 disconnect net/x25: Fix null-ptr-deref in x25_disconnect sh: Fix validation of system call number net: lan78xx: add missing endpoint sanity check net: lan78xx: fix transfer-buffer memory leak mlx4: disable device on shutdown mlxsw: core: Increase scope of RCU read-side critical section mlxsw: core: Free EMAD transactions using kfree_rcu() ibmvnic: Fix IRQ mapping disposal in error path mac80211: mesh: Free ie data when leaving mesh mac80211: mesh: Free pending skb when destroying a mpath arm64: csum: Fix handling of bad packets usb: hso: Fix debug compile warning on sparc32 qed: Disable "MFW indication via attention" SPAM every 5 minutes nfc: s3fwrn5: add missing release on skb in s3fwrn5_recv_frame parisc: add support for cmpxchg on u8 pointers net: ethernet: ravb: exit if re-initialization fails in tx timeout Revert "i2c: cadence: Fix the hold bit setting" xen-netfront: fix potential deadlock in xennet_remove() KVM: LAPIC: Prevent setting the tscdeadline timer if the lapic is hw disabled x86/i8259: Use printk_deferred() to prevent deadlock random32: update the net random state on interrupt and activity ARM: percpu.h: fix build error random: fix circular include dependency on arm64 after addition of percpu.h random32: remove net_rand_state from the latent entropy gcc plugin random32: move the pseudo-random 32-bit definitions to prandom.h ext4: fix direct I/O read error USB: serial: qcserial: add EM7305 QDL product ID net/mlx5e: Don't support phys switch id if not in switchdev mode ALSA: seq: oss: Serialize ioctls Bluetooth: Fix slab-out-of-bounds read in hci_extended_inquiry_result_evt() Bluetooth: Prevent out-of-bounds read in hci_inquiry_result_evt() Bluetooth: Prevent out-of-bounds read in hci_inquiry_result_with_rssi_evt() omapfb: dss: Fix max fclk divider for omap36xx vgacon: Fix for missing check in scrollback handling mtd: properly check all write ioctls for permissions leds: wm831x-status: fix use-after-free on unbind leds: da903x: fix use-after-free on unbind leds: lm3533: fix use-after-free on unbind leds: 88pm860x: fix use-after-free on unbind net/9p: validate fds in p9_fd_open drm/nouveau/fbcon: fix module unload when fbcon init has failed for some reason cfg80211: check vendor command doit pointer before use igb: reinit_locked() should be called with rtnl_lock atm: fix atm_dev refcnt leaks in atmtcp_remove_persistent tools lib traceevent: Fix memory leak in process_dynamic_array_len xattr: break delegations in {set,remove}xattr binder: Prevent context manager from incrementing ref 0 ipv4: Silence suspicious RCU usage warning ipv6: fix memory leaks on IPV6_ADDRFORM path vxlan: Ensure FDB dump is performed under RCU net: lan78xx: replace bogus endpoint lookup Revert "vxlan: fix tos value before xmit" usb: hso: check for return value in hso_serial_common_create() Smack: fix use-after-free in smk_write_relabel_self() tracepoint: Mark __tracepoint_string's __used gpio: fix oops resulting from calling of_get_named_gpio(NULL, ...) cgroup: add missing skcd->no_refcnt check in cgroup_sk_clone() EDAC: Fix reference count leaks arm64: dts: qcom: msm8916: Replace invalid bias-pull-none property arm64: dts: exynos: Fix silent hang after boot on Espresso m68k: mac: Don't send IOP message until channel is idle m68k: mac: Fix IOP status/control register writes platform/x86: intel-hid: Fix return value check in check_acpi_dev() platform/x86: intel-vbtn: Fix return value check in check_acpi_dev() ARM: at91: pm: add missing put_device() call in at91_pm_sram_init() ARM: socfpga: PM: add missing put_device() call in socfpga_setup_ocram_self_refresh() drm/tilcdc: fix leak & null ref in panel_connector_get_modes Bluetooth: add a mutex lock to avoid UAF in do_enale_set fs/btrfs: Add cond_resched() for try_release_extent_mapping() stalls drm/radeon: Fix reference count leaks caused by pm_runtime_get_sync video: fbdev: neofb: fix memory leak in neo_scan_monitor() md-cluster: fix wild pointer of unlock_all_bitmaps() drm/nouveau: fix multiple instances of reference count leaks drm/debugfs: fix plain echo to connector "force" attribute mm/mmap.c: Add cond_resched() for exit_mmap() CPU stalls brcmfmac: To fix Bss Info flag definition Bug iwlegacy: Check the return value of pcie_capability_read_*() usb: gadget: net2280: fix memory leak on probe error handling paths bdc: Fix bug causing crash after multiple disconnects dyndbg: fix a BUG_ON in ddebug_describe_flags bcache: fix super block seq numbers comparision in register_cache_set() ACPICA: Do not increment operation_region reference counts for field units agp/intel: Fix a memory leak on module initialisation failure video: fbdev: sm712fb: fix an issue about iounmap for a wrong address console: newport_con: fix an issue about leak related system resources video: pxafb: Fix the function used to balance a 'dma_alloc_coherent()' call iio: improve IIO_CONCENTRATION channel type description leds: lm355x: avoid enum conversion warning media: omap3isp: Add missed v4l2_ctrl_handler_free() for preview_init_entities() scsi: cumana_2: Fix different dev_id between request_irq() and free_irq() drm/mipi: use dcs write for mipi_dsi_dcs_set_tear_scanline cxl: Fix kobject memleak drm/radeon: fix array out-of-bounds read and write issues scsi: powertec: Fix different dev_id between request_irq() and free_irq() scsi: eesox: Fix different dev_id between request_irq() and free_irq() media: firewire: Using uninitialized values in node_probe() media: exynos4-is: Add missed check for pinctrl_lookup_state() xfs: fix reflink quota reservation accounting error PCI: Fix pci_cfg_wait queue locking problem leds: core: Flush scheduled work for system suspend drm: panel: simple: Fix bpc for LG LB070WV8 panel scsi: scsi_debug: Add check for sdebug_max_queue during module init mwifiex: Prevent memory corruption handling keys powerpc/vdso: Fix vdso cpu truncation staging: rtl8192u: fix a dubious looking mask before a shift PCI/ASPM: Add missing newline in sysfs 'policy' drm/imx: tve: fix regulator_disable error path USB: serial: iuu_phoenix: fix led-activity helpers usb: dwc2: Fix error path in gadget registration scsi: mesh: Fix panic after host or bus reset Smack: fix another vsscanf out of bounds Smack: prevent underflow in smk_set_cipso() power: supply: check if calc_soc succeeded in pm860x_init_battery selftests/powerpc: Fix CPU affinity for child process selftests/powerpc: Fix online CPU selection s390/qeth: don't process empty bridge port events wl1251: fix always return 0 error net: spider_net: Fix the size used in a 'dma_free_coherent()' call fsl/fman: use 32-bit unsigned integer fsl/fman: fix dereference null return value fsl/fman: fix unreachable code fsl/fman: check dereferencing null pointer fsl/fman: fix eth hash table allocation dlm: Fix kobject memleak pinctrl-single: fix pcs_parse_pinconf() return value drivers/net/wan/lapbether: Added needed_headroom and a skb->len check net/nfc/rawsock.c: add CAP_NET_RAW check. net: Set fput_needed iff FDPUT_FPUT is set USB: serial: cp210x: re-enable auto-RTS on open USB: serial: cp210x: enable usb generic throttle/unthrottle ALSA: usb-audio: Creative USB X-Fi Pro SB1095 volume knob support ALSA: usb-audio: fix overeager device match for MacroSilicon MS2109 ALSA: usb-audio: add quirk for Pioneer DDJ-RB crypto: qat - fix double free in qat_uclo_create_batch_init_list crypto: ccp - Fix use of merged scatterlists fs/minix: check return value of sb_getblk() fs/minix: don't allow getting deleted inodes fs/minix: reject too-large maximum file size ALSA: usb-audio: work around streaming quirk for MacroSilicon MS2109 9p: Fix memory leak in v9fs_mount parisc: mask out enable and reserved bits from sba imask ARM: 8992/1: Fix unwind_frame for clang-built kernels xen/balloon: fix accounting in alloc_xenballooned_pages error path xen/balloon: make the balloon wait interruptible smb3: warn on confusing error scenario with sec=krb5 PCI: hotplug: ACPI: Fix context refcounting in acpiphp_grab_context() btrfs: don't allocate anonymous block device for user invisible roots btrfs: only search for left_info if there is no right_info in try_merge_free_space btrfs: fix memory leaks after failure to lookup checksums during inode logging iio: dac: ad5592r: fix unbalanced mutex unlocks in ad5592r_read_raw() xtensa: fix xtensa_pmu_setup prototype powerpc: Fix circular dependency between percpu.h and mmu.h net: ethernet: stmmac: Disable hardware multicast filter net: stmmac: dwmac1000: provide multicast filter fallback net/compat: Add missing sock updates for SCM_RIGHTS md/raid5: Fix Force reconstruct-write io stuck in degraded raid5 bcache: allocate meta data pages as compound pages mac80211: fix misplaced while instead of if MIPS: CPU#0 is not hotpluggable ext2: fix missing percpu_counter_inc ocfs2: change slot number type s16 to u16 ftrace: Setup correct FTRACE_FL_REGS flags for module kprobes: Fix NULL pointer dereference at kprobe_ftrace_handler watchdog: f71808e_wdt: indicate WDIOF_CARDRESET support in watchdog_info.options watchdog: f71808e_wdt: remove use of wrong watchdog_info option watchdog: f71808e_wdt: clear watchdog timeout occurred flag pseries: Fix 64 bit logical memory block panic mfd: arizona: Ensure 32k clock is put on driver unbind and error USB: serial: ftdi_sio: make process-packet buffer unsigned USB: serial: ftdi_sio: clean up receive processing gpu: ipu-v3: image-convert: Combine rotate/no-rotate irq handlers iommu/omap: Check for failure of a call to omap_iommu_dump_ctx iommu/vt-d: Enforce PASID devTLB field mask i2c: rcar: slave: only send STOP event when we have been addressed clk: clk-atlas6: fix return value check in atlas6_clk_init() pwm: bcm-iproc: handle clk_get_rate() return Input: sentelic - fix error return when fsp_reg_write fails drm/vmwgfx: Fix two list_for_each loop exit tests net: qcom/emac: add missed clk_disable_unprepare in error path of emac_clks_phase1_init nfs: Fix getxattr kernel panic and memory overflow fs/ufs: avoid potential u32 multiplication overflow mfd: dln2: Run event handler loop under spinlock ALSA: echoaudio: Fix potential Oops in snd_echo_resume() sh: landisk: Add missing initialization of sh_io_port_base khugepaged: retract_page_tables() remember to test exit mm: Avoid calling build_all_zonelists_init under hotplug context Linux 4.9.233 Signed-off-by: Greg Kroah-Hartman <gregkh@google.com> Change-Id: Ied62cb0768f5bd8e989d75e7c2ccf6f1e6f2efd4
447 lines
12 KiB
C
447 lines
12 KiB
C
/*
|
|
* Xtensa Performance Monitor Module driver
|
|
* See Tensilica Debug User's Guide for PMU registers documentation.
|
|
*
|
|
* Copyright (C) 2015 Cadence Design Systems Inc.
|
|
*
|
|
* This program is free software; you can redistribute it and/or modify
|
|
* it under the terms of the GNU General Public License version 2 as
|
|
* published by the Free Software Foundation.
|
|
*/
|
|
|
|
#include <linux/interrupt.h>
|
|
#include <linux/irqdomain.h>
|
|
#include <linux/module.h>
|
|
#include <linux/of.h>
|
|
#include <linux/perf_event.h>
|
|
#include <linux/platform_device.h>
|
|
|
|
#include <asm/processor.h>
|
|
#include <asm/stacktrace.h>
|
|
|
|
/* Global control/status for all perf counters */
|
|
#define XTENSA_PMU_PMG 0x1000
|
|
/* Perf counter values */
|
|
#define XTENSA_PMU_PM(i) (0x1080 + (i) * 4)
|
|
/* Perf counter control registers */
|
|
#define XTENSA_PMU_PMCTRL(i) (0x1100 + (i) * 4)
|
|
/* Perf counter status registers */
|
|
#define XTENSA_PMU_PMSTAT(i) (0x1180 + (i) * 4)
|
|
|
|
#define XTENSA_PMU_PMG_PMEN 0x1
|
|
|
|
#define XTENSA_PMU_COUNTER_MASK 0xffffffffULL
|
|
#define XTENSA_PMU_COUNTER_MAX 0x7fffffff
|
|
|
|
#define XTENSA_PMU_PMCTRL_INTEN 0x00000001
|
|
#define XTENSA_PMU_PMCTRL_KRNLCNT 0x00000008
|
|
#define XTENSA_PMU_PMCTRL_TRACELEVEL 0x000000f0
|
|
#define XTENSA_PMU_PMCTRL_SELECT_SHIFT 8
|
|
#define XTENSA_PMU_PMCTRL_SELECT 0x00001f00
|
|
#define XTENSA_PMU_PMCTRL_MASK_SHIFT 16
|
|
#define XTENSA_PMU_PMCTRL_MASK 0xffff0000
|
|
|
|
#define XTENSA_PMU_MASK(select, mask) \
|
|
(((select) << XTENSA_PMU_PMCTRL_SELECT_SHIFT) | \
|
|
((mask) << XTENSA_PMU_PMCTRL_MASK_SHIFT) | \
|
|
XTENSA_PMU_PMCTRL_TRACELEVEL | \
|
|
XTENSA_PMU_PMCTRL_INTEN)
|
|
|
|
#define XTENSA_PMU_PMSTAT_OVFL 0x00000001
|
|
#define XTENSA_PMU_PMSTAT_INTASRT 0x00000010
|
|
|
|
struct xtensa_pmu_events {
|
|
/* Array of events currently on this core */
|
|
struct perf_event *event[XCHAL_NUM_PERF_COUNTERS];
|
|
/* Bitmap of used hardware counters */
|
|
unsigned long used_mask[BITS_TO_LONGS(XCHAL_NUM_PERF_COUNTERS)];
|
|
};
|
|
static DEFINE_PER_CPU(struct xtensa_pmu_events, xtensa_pmu_events);
|
|
|
|
static const u32 xtensa_hw_ctl[] = {
|
|
[PERF_COUNT_HW_CPU_CYCLES] = XTENSA_PMU_MASK(0, 0x1),
|
|
[PERF_COUNT_HW_INSTRUCTIONS] = XTENSA_PMU_MASK(2, 0xffff),
|
|
[PERF_COUNT_HW_CACHE_REFERENCES] = XTENSA_PMU_MASK(10, 0x1),
|
|
[PERF_COUNT_HW_CACHE_MISSES] = XTENSA_PMU_MASK(12, 0x1),
|
|
/* Taken and non-taken branches + taken loop ends */
|
|
[PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = XTENSA_PMU_MASK(2, 0x490),
|
|
/* Instruction-related + other global stall cycles */
|
|
[PERF_COUNT_HW_STALLED_CYCLES_FRONTEND] = XTENSA_PMU_MASK(4, 0x1ff),
|
|
/* Data-related global stall cycles */
|
|
[PERF_COUNT_HW_STALLED_CYCLES_BACKEND] = XTENSA_PMU_MASK(3, 0x1ff),
|
|
};
|
|
|
|
#define C(_x) PERF_COUNT_HW_CACHE_##_x
|
|
|
|
static const u32 xtensa_cache_ctl[][C(OP_MAX)][C(RESULT_MAX)] = {
|
|
[C(L1D)] = {
|
|
[C(OP_READ)] = {
|
|
[C(RESULT_ACCESS)] = XTENSA_PMU_MASK(10, 0x1),
|
|
[C(RESULT_MISS)] = XTENSA_PMU_MASK(10, 0x2),
|
|
},
|
|
[C(OP_WRITE)] = {
|
|
[C(RESULT_ACCESS)] = XTENSA_PMU_MASK(11, 0x1),
|
|
[C(RESULT_MISS)] = XTENSA_PMU_MASK(11, 0x2),
|
|
},
|
|
},
|
|
[C(L1I)] = {
|
|
[C(OP_READ)] = {
|
|
[C(RESULT_ACCESS)] = XTENSA_PMU_MASK(8, 0x1),
|
|
[C(RESULT_MISS)] = XTENSA_PMU_MASK(8, 0x2),
|
|
},
|
|
},
|
|
[C(DTLB)] = {
|
|
[C(OP_READ)] = {
|
|
[C(RESULT_ACCESS)] = XTENSA_PMU_MASK(9, 0x1),
|
|
[C(RESULT_MISS)] = XTENSA_PMU_MASK(9, 0x8),
|
|
},
|
|
},
|
|
[C(ITLB)] = {
|
|
[C(OP_READ)] = {
|
|
[C(RESULT_ACCESS)] = XTENSA_PMU_MASK(7, 0x1),
|
|
[C(RESULT_MISS)] = XTENSA_PMU_MASK(7, 0x8),
|
|
},
|
|
},
|
|
};
|
|
|
|
static int xtensa_pmu_cache_event(u64 config)
|
|
{
|
|
unsigned int cache_type, cache_op, cache_result;
|
|
int ret;
|
|
|
|
cache_type = (config >> 0) & 0xff;
|
|
cache_op = (config >> 8) & 0xff;
|
|
cache_result = (config >> 16) & 0xff;
|
|
|
|
if (cache_type >= ARRAY_SIZE(xtensa_cache_ctl) ||
|
|
cache_op >= C(OP_MAX) ||
|
|
cache_result >= C(RESULT_MAX))
|
|
return -EINVAL;
|
|
|
|
ret = xtensa_cache_ctl[cache_type][cache_op][cache_result];
|
|
|
|
if (ret == 0)
|
|
return -EINVAL;
|
|
|
|
return ret;
|
|
}
|
|
|
|
static inline uint32_t xtensa_pmu_read_counter(int idx)
|
|
{
|
|
return get_er(XTENSA_PMU_PM(idx));
|
|
}
|
|
|
|
static inline void xtensa_pmu_write_counter(int idx, uint32_t v)
|
|
{
|
|
set_er(v, XTENSA_PMU_PM(idx));
|
|
}
|
|
|
|
static void xtensa_perf_event_update(struct perf_event *event,
|
|
struct hw_perf_event *hwc, int idx)
|
|
{
|
|
uint64_t prev_raw_count, new_raw_count;
|
|
int64_t delta;
|
|
|
|
do {
|
|
prev_raw_count = local64_read(&hwc->prev_count);
|
|
new_raw_count = xtensa_pmu_read_counter(event->hw.idx);
|
|
} while (local64_cmpxchg(&hwc->prev_count, prev_raw_count,
|
|
new_raw_count) != prev_raw_count);
|
|
|
|
delta = (new_raw_count - prev_raw_count) & XTENSA_PMU_COUNTER_MASK;
|
|
|
|
local64_add(delta, &event->count);
|
|
local64_sub(delta, &hwc->period_left);
|
|
}
|
|
|
|
static bool xtensa_perf_event_set_period(struct perf_event *event,
|
|
struct hw_perf_event *hwc, int idx)
|
|
{
|
|
bool rc = false;
|
|
s64 left;
|
|
|
|
if (!is_sampling_event(event)) {
|
|
left = XTENSA_PMU_COUNTER_MAX;
|
|
} else {
|
|
s64 period = hwc->sample_period;
|
|
|
|
left = local64_read(&hwc->period_left);
|
|
if (left <= -period) {
|
|
left = period;
|
|
local64_set(&hwc->period_left, left);
|
|
hwc->last_period = period;
|
|
rc = true;
|
|
} else if (left <= 0) {
|
|
left += period;
|
|
local64_set(&hwc->period_left, left);
|
|
hwc->last_period = period;
|
|
rc = true;
|
|
}
|
|
if (left > XTENSA_PMU_COUNTER_MAX)
|
|
left = XTENSA_PMU_COUNTER_MAX;
|
|
}
|
|
|
|
local64_set(&hwc->prev_count, -left);
|
|
xtensa_pmu_write_counter(idx, -left);
|
|
perf_event_update_userpage(event);
|
|
|
|
return rc;
|
|
}
|
|
|
|
static void xtensa_pmu_enable(struct pmu *pmu)
|
|
{
|
|
set_er(get_er(XTENSA_PMU_PMG) | XTENSA_PMU_PMG_PMEN, XTENSA_PMU_PMG);
|
|
}
|
|
|
|
static void xtensa_pmu_disable(struct pmu *pmu)
|
|
{
|
|
set_er(get_er(XTENSA_PMU_PMG) & ~XTENSA_PMU_PMG_PMEN, XTENSA_PMU_PMG);
|
|
}
|
|
|
|
static int xtensa_pmu_event_init(struct perf_event *event)
|
|
{
|
|
int ret;
|
|
|
|
switch (event->attr.type) {
|
|
case PERF_TYPE_HARDWARE:
|
|
if (event->attr.config >= ARRAY_SIZE(xtensa_hw_ctl) ||
|
|
xtensa_hw_ctl[event->attr.config] == 0)
|
|
return -EINVAL;
|
|
event->hw.config = xtensa_hw_ctl[event->attr.config];
|
|
return 0;
|
|
|
|
case PERF_TYPE_HW_CACHE:
|
|
ret = xtensa_pmu_cache_event(event->attr.config);
|
|
if (ret < 0)
|
|
return ret;
|
|
event->hw.config = ret;
|
|
return 0;
|
|
|
|
case PERF_TYPE_RAW:
|
|
/* Not 'previous counter' select */
|
|
if ((event->attr.config & XTENSA_PMU_PMCTRL_SELECT) ==
|
|
(1 << XTENSA_PMU_PMCTRL_SELECT_SHIFT))
|
|
return -EINVAL;
|
|
event->hw.config = (event->attr.config &
|
|
(XTENSA_PMU_PMCTRL_KRNLCNT |
|
|
XTENSA_PMU_PMCTRL_TRACELEVEL |
|
|
XTENSA_PMU_PMCTRL_SELECT |
|
|
XTENSA_PMU_PMCTRL_MASK)) |
|
|
XTENSA_PMU_PMCTRL_INTEN;
|
|
return 0;
|
|
|
|
default:
|
|
return -ENOENT;
|
|
}
|
|
}
|
|
|
|
/*
|
|
* Starts/Stops a counter present on the PMU. The PMI handler
|
|
* should stop the counter when perf_event_overflow() returns
|
|
* !0. ->start() will be used to continue.
|
|
*/
|
|
static void xtensa_pmu_start(struct perf_event *event, int flags)
|
|
{
|
|
struct hw_perf_event *hwc = &event->hw;
|
|
int idx = hwc->idx;
|
|
|
|
if (WARN_ON_ONCE(idx == -1))
|
|
return;
|
|
|
|
if (flags & PERF_EF_RELOAD) {
|
|
WARN_ON_ONCE(!(event->hw.state & PERF_HES_UPTODATE));
|
|
xtensa_perf_event_set_period(event, hwc, idx);
|
|
}
|
|
|
|
hwc->state = 0;
|
|
|
|
set_er(hwc->config, XTENSA_PMU_PMCTRL(idx));
|
|
}
|
|
|
|
static void xtensa_pmu_stop(struct perf_event *event, int flags)
|
|
{
|
|
struct hw_perf_event *hwc = &event->hw;
|
|
int idx = hwc->idx;
|
|
|
|
if (!(hwc->state & PERF_HES_STOPPED)) {
|
|
set_er(0, XTENSA_PMU_PMCTRL(idx));
|
|
set_er(get_er(XTENSA_PMU_PMSTAT(idx)),
|
|
XTENSA_PMU_PMSTAT(idx));
|
|
hwc->state |= PERF_HES_STOPPED;
|
|
}
|
|
|
|
if ((flags & PERF_EF_UPDATE) &&
|
|
!(event->hw.state & PERF_HES_UPTODATE)) {
|
|
xtensa_perf_event_update(event, &event->hw, idx);
|
|
event->hw.state |= PERF_HES_UPTODATE;
|
|
}
|
|
}
|
|
|
|
/*
|
|
* Adds/Removes a counter to/from the PMU, can be done inside
|
|
* a transaction, see the ->*_txn() methods.
|
|
*/
|
|
static int xtensa_pmu_add(struct perf_event *event, int flags)
|
|
{
|
|
struct xtensa_pmu_events *ev = this_cpu_ptr(&xtensa_pmu_events);
|
|
struct hw_perf_event *hwc = &event->hw;
|
|
int idx = hwc->idx;
|
|
|
|
if (__test_and_set_bit(idx, ev->used_mask)) {
|
|
idx = find_first_zero_bit(ev->used_mask,
|
|
XCHAL_NUM_PERF_COUNTERS);
|
|
if (idx == XCHAL_NUM_PERF_COUNTERS)
|
|
return -EAGAIN;
|
|
|
|
__set_bit(idx, ev->used_mask);
|
|
hwc->idx = idx;
|
|
}
|
|
ev->event[idx] = event;
|
|
|
|
hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED;
|
|
|
|
if (flags & PERF_EF_START)
|
|
xtensa_pmu_start(event, PERF_EF_RELOAD);
|
|
|
|
perf_event_update_userpage(event);
|
|
return 0;
|
|
}
|
|
|
|
static void xtensa_pmu_del(struct perf_event *event, int flags)
|
|
{
|
|
struct xtensa_pmu_events *ev = this_cpu_ptr(&xtensa_pmu_events);
|
|
|
|
xtensa_pmu_stop(event, PERF_EF_UPDATE);
|
|
__clear_bit(event->hw.idx, ev->used_mask);
|
|
perf_event_update_userpage(event);
|
|
}
|
|
|
|
static void xtensa_pmu_read(struct perf_event *event)
|
|
{
|
|
xtensa_perf_event_update(event, &event->hw, event->hw.idx);
|
|
}
|
|
|
|
static int callchain_trace(struct stackframe *frame, void *data)
|
|
{
|
|
struct perf_callchain_entry_ctx *entry = data;
|
|
|
|
perf_callchain_store(entry, frame->pc);
|
|
return 0;
|
|
}
|
|
|
|
void perf_callchain_kernel(struct perf_callchain_entry_ctx *entry,
|
|
struct pt_regs *regs)
|
|
{
|
|
xtensa_backtrace_kernel(regs, entry->max_stack,
|
|
callchain_trace, NULL, entry);
|
|
}
|
|
|
|
void perf_callchain_user(struct perf_callchain_entry_ctx *entry,
|
|
struct pt_regs *regs)
|
|
{
|
|
xtensa_backtrace_user(regs, entry->max_stack,
|
|
callchain_trace, entry);
|
|
}
|
|
|
|
void perf_event_print_debug(void)
|
|
{
|
|
unsigned long flags;
|
|
unsigned i;
|
|
|
|
local_irq_save(flags);
|
|
pr_info("CPU#%d: PMG: 0x%08lx\n", smp_processor_id(),
|
|
get_er(XTENSA_PMU_PMG));
|
|
for (i = 0; i < XCHAL_NUM_PERF_COUNTERS; ++i)
|
|
pr_info("PM%d: 0x%08lx, PMCTRL%d: 0x%08lx, PMSTAT%d: 0x%08lx\n",
|
|
i, get_er(XTENSA_PMU_PM(i)),
|
|
i, get_er(XTENSA_PMU_PMCTRL(i)),
|
|
i, get_er(XTENSA_PMU_PMSTAT(i)));
|
|
local_irq_restore(flags);
|
|
}
|
|
|
|
irqreturn_t xtensa_pmu_irq_handler(int irq, void *dev_id)
|
|
{
|
|
irqreturn_t rc = IRQ_NONE;
|
|
struct xtensa_pmu_events *ev = this_cpu_ptr(&xtensa_pmu_events);
|
|
unsigned i;
|
|
|
|
for (i = find_first_bit(ev->used_mask, XCHAL_NUM_PERF_COUNTERS);
|
|
i < XCHAL_NUM_PERF_COUNTERS;
|
|
i = find_next_bit(ev->used_mask, XCHAL_NUM_PERF_COUNTERS, i + 1)) {
|
|
uint32_t v = get_er(XTENSA_PMU_PMSTAT(i));
|
|
struct perf_event *event = ev->event[i];
|
|
struct hw_perf_event *hwc = &event->hw;
|
|
u64 last_period;
|
|
|
|
if (!(v & XTENSA_PMU_PMSTAT_OVFL))
|
|
continue;
|
|
|
|
set_er(v, XTENSA_PMU_PMSTAT(i));
|
|
xtensa_perf_event_update(event, hwc, i);
|
|
last_period = hwc->last_period;
|
|
if (xtensa_perf_event_set_period(event, hwc, i)) {
|
|
struct perf_sample_data data;
|
|
struct pt_regs *regs = get_irq_regs();
|
|
|
|
perf_sample_data_init(&data, 0, last_period);
|
|
if (perf_event_overflow(event, &data, regs))
|
|
xtensa_pmu_stop(event, 0);
|
|
}
|
|
|
|
rc = IRQ_HANDLED;
|
|
}
|
|
return rc;
|
|
}
|
|
|
|
static struct pmu xtensa_pmu = {
|
|
.pmu_enable = xtensa_pmu_enable,
|
|
.pmu_disable = xtensa_pmu_disable,
|
|
.event_init = xtensa_pmu_event_init,
|
|
.add = xtensa_pmu_add,
|
|
.del = xtensa_pmu_del,
|
|
.start = xtensa_pmu_start,
|
|
.stop = xtensa_pmu_stop,
|
|
.read = xtensa_pmu_read,
|
|
};
|
|
|
|
static int xtensa_pmu_setup(unsigned int cpu)
|
|
{
|
|
unsigned i;
|
|
|
|
set_er(0, XTENSA_PMU_PMG);
|
|
for (i = 0; i < XCHAL_NUM_PERF_COUNTERS; ++i) {
|
|
set_er(0, XTENSA_PMU_PMCTRL(i));
|
|
set_er(get_er(XTENSA_PMU_PMSTAT(i)), XTENSA_PMU_PMSTAT(i));
|
|
}
|
|
return 0;
|
|
}
|
|
|
|
static int __init xtensa_pmu_init(void)
|
|
{
|
|
int ret;
|
|
int irq = irq_create_mapping(NULL, XCHAL_PROFILING_INTERRUPT);
|
|
|
|
ret = cpuhp_setup_state(CPUHP_AP_PERF_XTENSA_STARTING,
|
|
"AP_PERF_XTENSA_STARTING", xtensa_pmu_setup,
|
|
NULL);
|
|
if (ret) {
|
|
pr_err("xtensa_pmu: failed to register CPU-hotplug.\n");
|
|
return ret;
|
|
}
|
|
#if XTENSA_FAKE_NMI
|
|
enable_irq(irq);
|
|
#else
|
|
ret = request_irq(irq, xtensa_pmu_irq_handler, IRQF_PERCPU,
|
|
"pmu", NULL);
|
|
if (ret < 0)
|
|
return ret;
|
|
#endif
|
|
|
|
ret = perf_pmu_register(&xtensa_pmu, "cpu", PERF_TYPE_RAW);
|
|
if (ret)
|
|
free_irq(irq, NULL);
|
|
|
|
return ret;
|
|
}
|
|
early_initcall(xtensa_pmu_init);
|