perf/x86: Fixes for Nehalem-EX uncore driver

This patch includes following fixes and update:
 - Only some events in the Sbox and Mbox can use the match/mask
   registers, add code to check this.
 - The format definitions for xbr_mm_cfg and xbr_match registers
   in the Rbox are wrong, xbr_mm_cfg should use 32 bits, xbr_match
   should use 64 bits.
 - Cleanup the Rbox code. Compute the addresses extra registers in
   the enable_event function instead of the hw_config function.
   This simplifies the code in nhmex_rbox_alter_er().

Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1344229882-3907-2-git-send-email-zheng.z.yan@intel.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

2 files changed, 87 insertions, 109 deletions

diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.c b/arch/x86/kernel/cpu/perf_event_intel_uncore.c
index a7ccd68aa13a..84434e2a676f 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_uncore.c
+++ b/arch/x86/kernel/cpu/perf_event_intel_uncore.c
@@ -796,7 +796,6 @@ static struct intel_uncore_type *nhm_msr_uncores[] = {
 
 DEFINE_UNCORE_FORMAT_ATTR(event5, event, "config:1-5");
 DEFINE_UNCORE_FORMAT_ATTR(counter, counter, "config:6-7");
-DEFINE_UNCORE_FORMAT_ATTR(mm_cfg, mm_cfg, "config:63");
 DEFINE_UNCORE_FORMAT_ATTR(match, match, "config1:0-63");
 DEFINE_UNCORE_FORMAT_ATTR(mask, mask, "config2:0-63");
 
@@ -1032,24 +1031,22 @@ static struct intel_uncore_type nhmex_uncore_bbox = {
 
 static int nhmex_sbox_hw_config(struct intel_uncore_box *box, struct perf_event *event)
 {
-        struct hw_perf_event_extra *reg1 = &event->hw.extra_reg;
-        struct hw_perf_event_extra *reg2 = &event->hw.branch_reg;
+        struct hw_perf_event *hwc = &event->hw;
+        struct hw_perf_event_extra *reg1 = &hwc->extra_reg;
+        struct hw_perf_event_extra *reg2 = &hwc->branch_reg;
 
-        if (event->attr.config & NHMEX_S_PMON_MM_CFG_EN) {
-                reg1->config = event->attr.config1;
-                reg2->config = event->attr.config2;
-        } else {
-                reg1->config = ~0ULL;
-                reg2->config = ~0ULL;
-        }
+        /* only TO_R_PROG_EV event uses the match/mask register */
+        if ((hwc->config & NHMEX_PMON_CTL_EV_SEL_MASK) !=
+            NHMEX_S_EVENT_TO_R_PROG_EV)
+                return 0;
 
         if (box->pmu->pmu_idx == 0)
                 reg1->reg = NHMEX_S0_MSR_MM_CFG;
         else
                 reg1->reg = NHMEX_S1_MSR_MM_CFG;
-
         reg1->idx = 0;
-
+        reg1->config = event->attr.config1;
+        reg2->config = event->attr.config2;
         return 0;
 }
 
@@ -1059,8 +1056,8 @@ static void nhmex_sbox_msr_enable_event(struct intel_uncore_box *box, struct per
         struct hw_perf_event_extra *reg1 = &hwc->extra_reg;
         struct hw_perf_event_extra *reg2 = &hwc->branch_reg;
 
-        wrmsrl(reg1->reg, 0);
-        if (reg1->config != ~0ULL || reg2->config != ~0ULL) {
+        if (reg1->idx != EXTRA_REG_NONE) {
+                wrmsrl(reg1->reg, 0);
                 wrmsrl(reg1->reg + 1, reg1->config);
                 wrmsrl(reg1->reg + 2, reg2->config);
                 wrmsrl(reg1->reg, NHMEX_S_PMON_MM_CFG_EN);
@@ -1074,7 +1071,6 @@ static struct attribute *nhmex_uncore_sbox_formats_attr[] = {
         &format_attr_edge.attr,
         &format_attr_inv.attr,
         &format_attr_thresh8.attr,
-        &format_attr_mm_cfg.attr,
         &format_attr_match.attr,
         &format_attr_mask.attr,
         NULL,
@@ -1264,7 +1260,8 @@ again:
         }
 
         /* for the match/mask registers */
-        if ((uncore_box_is_fake(box) || !reg2->alloc) &&
+        if (reg2->idx != EXTRA_REG_NONE &&
+            (uncore_box_is_fake(box) || !reg2->alloc) &&
             !nhmex_mbox_get_shared_reg(box, reg2->idx, reg2->config))
                 goto fail;
 
@@ -1278,7 +1275,8 @@ again:
                 if (idx[0] != 0xff && idx[0] != __BITS_VALUE(reg1->idx, 0, 8))
                         nhmex_mbox_alter_er(event, idx[0], true);
                 reg1->alloc |= alloc;
-                reg2->alloc = 1;
+                if (reg2->idx != EXTRA_REG_NONE)
+                        reg2->alloc = 1;
         }
         return NULL;
 fail:
@@ -1342,9 +1340,6 @@ static int nhmex_mbox_hw_config(struct intel_uncore_box *box, struct perf_event
         struct extra_reg *er;
         unsigned msr;
         int reg_idx = 0;
-
-        if (WARN_ON_ONCE(reg1->idx != -1))
-                return -EINVAL;
         /*
          * The mbox events may require 2 extra MSRs at the most. But only
          * the lower 32 bits in these MSRs are significant, so we can use
@@ -1355,11 +1350,6 @@ static int nhmex_mbox_hw_config(struct intel_uncore_box *box, struct perf_event
                         continue;
                 if (event->attr.config1 & ~er->valid_mask)
                         return -EINVAL;
-                if (er->idx == __BITS_VALUE(reg1->idx, 0, 8) ||
-                    er->idx == __BITS_VALUE(reg1->idx, 1, 8))
-                        continue;
-                if (WARN_ON_ONCE(reg_idx >= 2))
-                        return -EINVAL;
 
                 msr = er->msr + type->msr_offset * box->pmu->pmu_idx;
                 if (WARN_ON_ONCE(msr >= 0xffff || er->idx >= 0xff))
@@ -1368,6 +1358,8 @@ static int nhmex_mbox_hw_config(struct intel_uncore_box *box, struct perf_event
                 /* always use the 32~63 bits to pass the PLD config */
                 if (er->idx == EXTRA_REG_NHMEX_M_PLD)
                         reg_idx = 1;
+                else if (WARN_ON_ONCE(reg_idx > 0))
+                        return -EINVAL;
 
                 reg1->idx &= ~(0xff << (reg_idx * 8));
                 reg1->reg &= ~(0xffff << (reg_idx * 16));
@@ -1376,17 +1368,21 @@ static int nhmex_mbox_hw_config(struct intel_uncore_box *box, struct perf_event
                 reg1->config = event->attr.config1;
                 reg_idx++;
         }
-        /* use config2 to pass the filter config */
-        reg2->idx = EXTRA_REG_NHMEX_M_FILTER;
-        if (event->attr.config2 & NHMEX_M_PMON_MM_CFG_EN)
-                reg2->config = event->attr.config2;
-        else
-                reg2->config = ~0ULL;
-        if (box->pmu->pmu_idx == 0)
-                reg2->reg = NHMEX_M0_MSR_PMU_MM_CFG;
-        else
-                reg2->reg = NHMEX_M1_MSR_PMU_MM_CFG;
-
+        /*
+         * The mbox only provides ability to perform address matching
+         * for the PLD events.
+         */
+        if (reg_idx == 2) {
+                reg2->idx = EXTRA_REG_NHMEX_M_FILTER;
+                if (event->attr.config2 & NHMEX_M_PMON_MM_CFG_EN)
+                        reg2->config = event->attr.config2;
+                else
+                        reg2->config = ~0ULL;
+                if (box->pmu->pmu_idx == 0)
+                        reg2->reg = NHMEX_M0_MSR_PMU_MM_CFG;
+                else
+                        reg2->reg = NHMEX_M1_MSR_PMU_MM_CFG;
+        }
         return 0;
 }
 
@@ -1422,34 +1418,36 @@ static void nhmex_mbox_msr_enable_event(struct intel_uncore_box *box, struct per
                 wrmsrl(__BITS_VALUE(reg1->reg, 1, 16),
                         nhmex_mbox_shared_reg_config(box, idx));
 
-        wrmsrl(reg2->reg, 0);
-        if (reg2->config != ~0ULL) {
-                wrmsrl(reg2->reg + 1,
-                        reg2->config & NHMEX_M_PMON_ADDR_MATCH_MASK);
-                wrmsrl(reg2->reg + 2, NHMEX_M_PMON_ADDR_MASK_MASK &
-                        (reg2->config >> NHMEX_M_PMON_ADDR_MASK_SHIFT));
-                wrmsrl(reg2->reg, NHMEX_M_PMON_MM_CFG_EN);
+        if (reg2->idx != EXTRA_REG_NONE) {
+                wrmsrl(reg2->reg, 0);
+                if (reg2->config != ~0ULL) {
+                        wrmsrl(reg2->reg + 1,
+                                reg2->config & NHMEX_M_PMON_ADDR_MATCH_MASK);
+                        wrmsrl(reg2->reg + 2, NHMEX_M_PMON_ADDR_MASK_MASK &
+                                (reg2->config >> NHMEX_M_PMON_ADDR_MASK_SHIFT));
+                        wrmsrl(reg2->reg, NHMEX_M_PMON_MM_CFG_EN);
+                }
         }
 
         wrmsrl(hwc->config_base, hwc->config | NHMEX_PMON_CTL_EN_BIT0);
 }
 
-DEFINE_UNCORE_FORMAT_ATTR(count_mode,   count_mode,     "config:2-3");
-DEFINE_UNCORE_FORMAT_ATTR(storage_mode, storage_mode,   "config:4-5");
-DEFINE_UNCORE_FORMAT_ATTR(wrap_mode,    wrap_mode,      "config:6");
-DEFINE_UNCORE_FORMAT_ATTR(flag_mode,    flag_mode,      "config:7");
-DEFINE_UNCORE_FORMAT_ATTR(inc_sel,      inc_sel,        "config:9-13");
-DEFINE_UNCORE_FORMAT_ATTR(set_flag_sel, set_flag_sel,   "config:19-21");
-DEFINE_UNCORE_FORMAT_ATTR(filter_cfg,   filter_cfg,     "config2:63");
-DEFINE_UNCORE_FORMAT_ATTR(filter_match, filter_match,   "config2:0-33");
-DEFINE_UNCORE_FORMAT_ATTR(filter_mask,  filter_mask,    "config2:34-61");
-DEFINE_UNCORE_FORMAT_ATTR(dsp,          dsp,            "config1:0-31");
-DEFINE_UNCORE_FORMAT_ATTR(thr,          thr,            "config1:0-31");
-DEFINE_UNCORE_FORMAT_ATTR(fvc,          fvc,            "config1:0-31");
-DEFINE_UNCORE_FORMAT_ATTR(pgt,          pgt,            "config1:0-31");
-DEFINE_UNCORE_FORMAT_ATTR(map,          map,            "config1:0-31");
-DEFINE_UNCORE_FORMAT_ATTR(iss,          iss,            "config1:0-31");
-DEFINE_UNCORE_FORMAT_ATTR(pld,          pld,            "config1:32-63");
+DEFINE_UNCORE_FORMAT_ATTR(count_mode,           count_mode,     "config:2-3");
+DEFINE_UNCORE_FORMAT_ATTR(storage_mode,         storage_mode,   "config:4-5");
+DEFINE_UNCORE_FORMAT_ATTR(wrap_mode,            wrap_mode,      "config:6");
+DEFINE_UNCORE_FORMAT_ATTR(flag_mode,            flag_mode,      "config:7");
+DEFINE_UNCORE_FORMAT_ATTR(inc_sel,              inc_sel,        "config:9-13");
+DEFINE_UNCORE_FORMAT_ATTR(set_flag_sel,         set_flag_sel,   "config:19-21");
+DEFINE_UNCORE_FORMAT_ATTR(filter_cfg_en,        filter_cfg_en,  "config2:63");
+DEFINE_UNCORE_FORMAT_ATTR(filter_match,         filter_match,   "config2:0-33");
+DEFINE_UNCORE_FORMAT_ATTR(filter_mask,          filter_mask,    "config2:34-61");
+DEFINE_UNCORE_FORMAT_ATTR(dsp,                  dsp,            "config1:0-31");
+DEFINE_UNCORE_FORMAT_ATTR(thr,                  thr,            "config1:0-31");
+DEFINE_UNCORE_FORMAT_ATTR(fvc,                  fvc,            "config1:0-31");
+DEFINE_UNCORE_FORMAT_ATTR(pgt,                  pgt,            "config1:0-31");
+DEFINE_UNCORE_FORMAT_ATTR(map,                  map,            "config1:0-31");
+DEFINE_UNCORE_FORMAT_ATTR(iss,                  iss,            "config1:0-31");
+DEFINE_UNCORE_FORMAT_ATTR(pld,                  pld,            "config1:32-63");
 
 static struct attribute *nhmex_uncore_mbox_formats_attr[] = {
         &format_attr_count_mode.attr,
@@ -1458,7 +1456,7 @@ static struct attribute *nhmex_uncore_mbox_formats_attr[] = {
         &format_attr_flag_mode.attr,
         &format_attr_inc_sel.attr,
         &format_attr_set_flag_sel.attr,
-        &format_attr_filter_cfg.attr,
+        &format_attr_filter_cfg_en.attr,
         &format_attr_filter_match.attr,
         &format_attr_filter_mask.attr,
         &format_attr_dsp.attr,
@@ -1513,7 +1511,7 @@ void nhmex_rbox_alter_er(struct intel_uncore_box *box, struct perf_event *event)
         struct hw_perf_event_extra *reg1 = &hwc->extra_reg;
         int port;
 
-        /* adjust the main event selector */
+        /* adjust the main event selector and extra register index */
         if (reg1->idx % 2) {
                 reg1->idx--;
                 hwc->config -= 1 << NHMEX_R_PMON_CTL_EV_SEL_SHIFT;
@@ -1522,29 +1520,17 @@ void nhmex_rbox_alter_er(struct intel_uncore_box *box, struct perf_event *event)
                 hwc->config += 1 << NHMEX_R_PMON_CTL_EV_SEL_SHIFT;
         }
 
-        /* adjust address or config of extra register */
+        /* adjust extra register config */
         port = reg1->idx / 6 + box->pmu->pmu_idx * 4;
         switch (reg1->idx % 6) {
-        case 0:
-                reg1->reg = NHMEX_R_MSR_PORTN_IPERF_CFG0(port);
-                break;
-        case 1:
-                reg1->reg = NHMEX_R_MSR_PORTN_IPERF_CFG1(port);
-                break;
         case 2:
-                /* the 8~15 bits to the 0~7 bits */
+                /* shift the 8~15 bits to the 0~7 bits */
                 reg1->config >>= 8;
                 break;
         case 3:
-                /* the 0~7 bits to the 8~15 bits */
+                /* shift the 0~7 bits to the 8~15 bits */
                 reg1->config <<= 8;
                 break;
-        case 4:
-                reg1->reg = NHMEX_R_MSR_PORTN_XBR_SET1_MM_CFG(port);
-                break;
-        case 5:
-                reg1->reg = NHMEX_R_MSR_PORTN_XBR_SET2_MM_CFG(port);
-                break;
         };
 }
 
@@ -1671,7 +1657,7 @@ static int nhmex_rbox_hw_config(struct intel_uncore_box *box, struct perf_event
         struct hw_perf_event *hwc = &event->hw;
         struct hw_perf_event_extra *reg1 = &event->hw.extra_reg;
         struct hw_perf_event_extra *reg2 = &event->hw.branch_reg;
-        int port, idx;
+        int idx;
 
         idx = (event->hw.config & NHMEX_R_PMON_CTL_EV_SEL_MASK) >>
                 NHMEX_R_PMON_CTL_EV_SEL_SHIFT;
@@ -1681,27 +1667,11 @@ static int nhmex_rbox_hw_config(struct intel_uncore_box *box, struct perf_event
         reg1->idx = idx;
         reg1->config = event->attr.config1;
 
-        port = idx / 6 + box->pmu->pmu_idx * 4;
-        idx %= 6;
-        switch (idx) {
-        case 0:
-                reg1->reg = NHMEX_R_MSR_PORTN_IPERF_CFG0(port);
-                break;
-        case 1:
-                reg1->reg = NHMEX_R_MSR_PORTN_IPERF_CFG1(port);
-                break;
-        case 2:
-        case 3:
-                reg1->reg = NHMEX_R_MSR_PORTN_QLX_CFG(port);
-                break;
+        switch (idx % 6) {
         case 4:
         case 5:
-                if (idx == 4)
-                        reg1->reg = NHMEX_R_MSR_PORTN_XBR_SET1_MM_CFG(port);
-                else
-                        reg1->reg = NHMEX_R_MSR_PORTN_XBR_SET2_MM_CFG(port);
-                reg2->config = event->attr.config2;
                 hwc->config |= event->attr.config & (~0ULL << 32);
+                reg2->config = event->attr.config2;
                 break;
         };
         return 0;
@@ -1727,28 +1697,34 @@ static void nhmex_rbox_msr_enable_event(struct intel_uncore_box *box, struct per
         struct hw_perf_event *hwc = &event->hw;
         struct hw_perf_event_extra *reg1 = &hwc->extra_reg;
         struct hw_perf_event_extra *reg2 = &hwc->branch_reg;
-        int idx, er_idx;
+        int idx, port;
 
-        idx = reg1->idx % 6;
-        er_idx = idx;
-        if (er_idx > 2)
-                er_idx--;
-        er_idx += (reg1->idx / 6) * 5;
+        idx = reg1->idx;
+        port = idx / 6 + box->pmu->pmu_idx * 4;
 
-        switch (idx) {
+        switch (idx % 6) {
         case 0:
+                wrmsrl(NHMEX_R_MSR_PORTN_IPERF_CFG0(port), reg1->config);
+                break;
         case 1:
-                wrmsrl(reg1->reg, reg1->config);
+                wrmsrl(NHMEX_R_MSR_PORTN_IPERF_CFG1(port), reg1->config);
                 break;
         case 2:
         case 3:
-                wrmsrl(reg1->reg, nhmex_rbox_shared_reg_config(box, er_idx));
+                wrmsrl(NHMEX_R_MSR_PORTN_QLX_CFG(port),
+                        nhmex_rbox_shared_reg_config(box, 2 + (idx / 6) * 5));
                 break;
         case 4:
+                wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET1_MM_CFG(port),
+                        hwc->config >> 32);
+                wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET1_MATCH(port), reg1->config);
+                wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET1_MASK(port), reg2->config);
+                break;
         case 5:
-                wrmsrl(reg1->reg, reg1->config);
-                wrmsrl(reg1->reg + 1, hwc->config >> 32);
-                wrmsrl(reg1->reg + 2, reg2->config);
+                wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET2_MM_CFG(port),
+                        hwc->config >> 32);
+                wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET2_MATCH(port), reg1->config);
+                wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET2_MASK(port), reg2->config);
                 break;
         };
 
@@ -1756,8 +1732,8 @@ static void nhmex_rbox_msr_enable_event(struct intel_uncore_box *box, struct per
                 (hwc->config & NHMEX_R_PMON_CTL_EV_SEL_MASK));
 }
 
-DEFINE_UNCORE_FORMAT_ATTR(xbr_match, xbr_match, "config:32-63");
-DEFINE_UNCORE_FORMAT_ATTR(xbr_mm_cfg, xbr_mm_cfg, "config1:0-63");
+DEFINE_UNCORE_FORMAT_ATTR(xbr_mm_cfg, xbr_mm_cfg, "config:32-63");
+DEFINE_UNCORE_FORMAT_ATTR(xbr_match, xbr_match, "config1:0-63");
 DEFINE_UNCORE_FORMAT_ATTR(xbr_mask, xbr_mask, "config2:0-63");
 DEFINE_UNCORE_FORMAT_ATTR(qlx_cfg, qlx_cfg, "config1:0-15");
 DEFINE_UNCORE_FORMAT_ATTR(iperf_cfg, iperf_cfg, "config1:0-31");
@@ -2303,6 +2279,7 @@ int uncore_pmu_event_init(struct perf_event *event)
         event->hw.idx = -1;
         event->hw.last_tag = ~0ULL;
         event->hw.extra_reg.idx = EXTRA_REG_NONE;
+        event->hw.branch_reg.idx = EXTRA_REG_NONE;
 
         if (event->attr.config == UNCORE_FIXED_EVENT) {
                 /* no fixed counter */
diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.h b/arch/x86/kernel/cpu/perf_event_intel_uncore.h
index c9e5dc56630a..8384e9b543b7 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_uncore.h
+++ b/arch/x86/kernel/cpu/perf_event_intel_uncore.h
@@ -230,6 +230,7 @@
 #define NHMEX_S1_MSR_MASK                       0xe5a
 
 #define NHMEX_S_PMON_MM_CFG_EN                  (0x1ULL << 63)
+#define NHMEX_S_EVENT_TO_R_PROG_EV              0
 
 /* NHM-EX Mbox */
 #define NHMEX_M0_MSR_GLOBAL_CTL                 0xca0