* [PATCH 03/15] powerpc/msi: Improve IRQ bitmap allocator
From: Michael Neuling @ 2014-09-18 8:26 UTC (permalink / raw)
To: greg, arnd, mpe, benh
Cc: cbe-oss-dev, mikey, imunsie, linux-kernel, linuxppc-dev, jk,
anton
In-Reply-To: <1411028820-29933-1-git-send-email-mikey@neuling.org>
From: Ian Munsie <imunsie@au1.ibm.com>
Currently msi_bitmap_alloc_hwirqs() will round up any IRQ allocation requests
to the nearest power of 2. eg. ask for 5 IRQs and you'll get 8. This wastes a
lot of IRQs which can be a scarce resource.
For cxl we can require multiple IRQs for every contexts that is attached to the
accelerator. For AFU directed accelerators, there may be 1000s of contexts
attached, hence we can easily run out of IRQs, especially if we are needlessly
wasting them.
This changes the msi_bitmap_alloc_hwirqs() to allocate only the required number
of IRQs, hence avoiding this wastage.
Signed-off-by: Ian Munsie <imunsie@au1.ibm.com>
Signed-off-by: Michael Neuling <mikey@neuling.org>
---
arch/powerpc/sysdev/msi_bitmap.c | 18 +++++++++++++-----
1 file changed, 13 insertions(+), 5 deletions(-)
diff --git a/arch/powerpc/sysdev/msi_bitmap.c b/arch/powerpc/sysdev/msi_bitmap.c
index 2ff6302..e001559 100644
--- a/arch/powerpc/sysdev/msi_bitmap.c
+++ b/arch/powerpc/sysdev/msi_bitmap.c
@@ -24,28 +24,36 @@ int msi_bitmap_alloc_hwirqs(struct msi_bitmap *bmp, int num)
* This is fast, but stricter than we need. We might want to add
* a fallback routine which does a linear search with no alignment.
*/
- offset = bitmap_find_free_region(bmp->bitmap, bmp->irq_count, order);
+ offset = bitmap_find_next_zero_area(bmp->bitmap, bmp->irq_count, 0,
+ num, (1 << order) - 1);
+ if (offset > bmp->irq_count)
+ goto err;
+ bitmap_set(bmp->bitmap, offset, num);
spin_unlock_irqrestore(&bmp->lock, flags);
pr_debug("msi_bitmap: allocated 0x%x (2^%d) at offset 0x%x\n",
num, order, offset);
return offset;
+err:
+ spin_unlock_irqrestore(&bmp->lock, flags);
+ return -ENOMEM;
}
+EXPORT_SYMBOL(msi_bitmap_alloc_hwirqs);
void msi_bitmap_free_hwirqs(struct msi_bitmap *bmp, unsigned int offset,
unsigned int num)
{
unsigned long flags;
- int order = get_count_order(num);
- pr_debug("msi_bitmap: freeing 0x%x (2^%d) at offset 0x%x\n",
- num, order, offset);
+ pr_debug("msi_bitmap: freeing 0x%x at offset 0x%x\n",
+ num, offset);
spin_lock_irqsave(&bmp->lock, flags);
- bitmap_release_region(bmp->bitmap, offset, order);
+ bitmap_clear(bmp->bitmap, offset, num);
spin_unlock_irqrestore(&bmp->lock, flags);
}
+EXPORT_SYMBOL(msi_bitmap_free_hwirqs);
void msi_bitmap_reserve_hwirq(struct msi_bitmap *bmp, unsigned int hwirq)
{
--
1.9.1
^ permalink raw reply related
* [PATCH 02/15] powerpc/cell: Move data segment faulting code out of cell platform
From: Michael Neuling @ 2014-09-18 8:26 UTC (permalink / raw)
To: greg, arnd, mpe, benh
Cc: cbe-oss-dev, mikey, imunsie, linux-kernel, linuxppc-dev, jk,
anton
In-Reply-To: <1411028820-29933-1-git-send-email-mikey@neuling.org>
From: Ian Munsie <imunsie@au1.ibm.com>
__spu_trap_data_seg() currently contains code to determine the VSID and ESID
required for a particular EA and mm struct.
This code is generically useful for other co-processors. This moves the code
of the cell platform so it can be used by other powerpc code.
Signed-off-by: Ian Munsie <imunsie@au1.ibm.com>
Signed-off-by: Michael Neuling <mikey@neuling.org>
---
arch/powerpc/include/asm/mmu-hash64.h | 2 ++
arch/powerpc/mm/copro_fault.c | 48 ++++++++++++++++++++++++++++++++++
arch/powerpc/mm/slb.c | 3 ---
arch/powerpc/platforms/cell/spu_base.c | 41 +++--------------------------
4 files changed, 54 insertions(+), 40 deletions(-)
diff --git a/arch/powerpc/include/asm/mmu-hash64.h b/arch/powerpc/include/asm/mmu-hash64.h
index d765144..fd19a53 100644
--- a/arch/powerpc/include/asm/mmu-hash64.h
+++ b/arch/powerpc/include/asm/mmu-hash64.h
@@ -180,6 +180,8 @@ static inline unsigned int mmu_psize_to_shift(unsigned int mmu_psize)
* we work in all cases including 4k page size.
*/
#define VPN_SHIFT 12
+#define slb_vsid_shift(ssize) \
+ ((ssize) == MMU_SEGSIZE_256M ? SLB_VSID_SHIFT : SLB_VSID_SHIFT_1T)
/*
* HPTE Large Page (LP) details
diff --git a/arch/powerpc/mm/copro_fault.c b/arch/powerpc/mm/copro_fault.c
index ba7df14..4105a63 100644
--- a/arch/powerpc/mm/copro_fault.c
+++ b/arch/powerpc/mm/copro_fault.c
@@ -90,3 +90,51 @@ out_unlock:
return ret;
}
EXPORT_SYMBOL_GPL(copro_handle_mm_fault);
+
+int copro_data_segment(struct mm_struct *mm, u64 ea, u64 *esid, u64 *vsid)
+{
+ int psize, ssize;
+
+ *esid = (ea & ESID_MASK) | SLB_ESID_V;
+
+ switch (REGION_ID(ea)) {
+ case USER_REGION_ID:
+ pr_devel("copro_data_segment: 0x%llx -- USER_REGION_ID\n", ea);
+#ifdef CONFIG_PPC_MM_SLICES
+ psize = get_slice_psize(mm, ea);
+#else
+ psize = mm->context.user_psize;
+#endif
+ ssize = user_segment_size(ea);
+ *vsid = (get_vsid(mm->context.id, ea, ssize)
+ << slb_vsid_shift(ssize)) | SLB_VSID_USER
+ | (ssize == MMU_SEGSIZE_1T ? SLB_VSID_B_1T : 0);
+ break;
+ case VMALLOC_REGION_ID:
+ pr_devel("copro_data_segment: 0x%llx -- VMALLOC_REGION_ID\n", ea);
+ if (ea < VMALLOC_END)
+ psize = mmu_vmalloc_psize;
+ else
+ psize = mmu_io_psize;
+ *vsid = (get_kernel_vsid(ea, mmu_kernel_ssize)
+ << SLB_VSID_SHIFT) | SLB_VSID_KERNEL
+ | (mmu_kernel_ssize == MMU_SEGSIZE_1T ? SLB_VSID_B_1T : 0);
+ break;
+ case KERNEL_REGION_ID:
+ pr_devel("copro_data_segment: 0x%llx -- KERNEL_REGION_ID\n", ea);
+ psize = mmu_linear_psize;
+ *vsid = (get_kernel_vsid(ea, mmu_kernel_ssize)
+ << SLB_VSID_SHIFT) | SLB_VSID_KERNEL
+ | (mmu_kernel_ssize == MMU_SEGSIZE_1T ? SLB_VSID_B_1T : 0);
+ break;
+ default:
+ /* Future: support kernel segments so that drivers can use the
+ * CoProcessors */
+ pr_debug("invalid region access at %016llx\n", ea);
+ return 1;
+ }
+ *vsid |= mmu_psize_defs[psize].sllp;
+
+ return 0;
+}
+EXPORT_SYMBOL_GPL(copro_data_segment);
diff --git a/arch/powerpc/mm/slb.c b/arch/powerpc/mm/slb.c
index 0399a67..6e450ca 100644
--- a/arch/powerpc/mm/slb.c
+++ b/arch/powerpc/mm/slb.c
@@ -46,9 +46,6 @@ static inline unsigned long mk_esid_data(unsigned long ea, int ssize,
return (ea & slb_esid_mask(ssize)) | SLB_ESID_V | slot;
}
-#define slb_vsid_shift(ssize) \
- ((ssize) == MMU_SEGSIZE_256M? SLB_VSID_SHIFT: SLB_VSID_SHIFT_1T)
-
static inline unsigned long mk_vsid_data(unsigned long ea, int ssize,
unsigned long flags)
{
diff --git a/arch/powerpc/platforms/cell/spu_base.c b/arch/powerpc/platforms/cell/spu_base.c
index 2930d1e..fe004b1 100644
--- a/arch/powerpc/platforms/cell/spu_base.c
+++ b/arch/powerpc/platforms/cell/spu_base.c
@@ -167,45 +167,12 @@ static inline void spu_load_slb(struct spu *spu, int slbe, struct spu_slb *slb)
static int __spu_trap_data_seg(struct spu *spu, unsigned long ea)
{
- struct mm_struct *mm = spu->mm;
struct spu_slb slb;
- int psize;
-
- pr_debug("%s\n", __func__);
-
- slb.esid = (ea & ESID_MASK) | SLB_ESID_V;
+ int ret;
- switch(REGION_ID(ea)) {
- case USER_REGION_ID:
-#ifdef CONFIG_PPC_MM_SLICES
- psize = get_slice_psize(mm, ea);
-#else
- psize = mm->context.user_psize;
-#endif
- slb.vsid = (get_vsid(mm->context.id, ea, MMU_SEGSIZE_256M)
- << SLB_VSID_SHIFT) | SLB_VSID_USER;
- break;
- case VMALLOC_REGION_ID:
- if (ea < VMALLOC_END)
- psize = mmu_vmalloc_psize;
- else
- psize = mmu_io_psize;
- slb.vsid = (get_kernel_vsid(ea, MMU_SEGSIZE_256M)
- << SLB_VSID_SHIFT) | SLB_VSID_KERNEL;
- break;
- case KERNEL_REGION_ID:
- psize = mmu_linear_psize;
- slb.vsid = (get_kernel_vsid(ea, MMU_SEGSIZE_256M)
- << SLB_VSID_SHIFT) | SLB_VSID_KERNEL;
- break;
- default:
- /* Future: support kernel segments so that drivers
- * can use SPUs.
- */
- pr_debug("invalid region access at %016lx\n", ea);
- return 1;
- }
- slb.vsid |= mmu_psize_defs[psize].sllp;
+ ret = copro_data_segment(spu->mm, ea, &slb.esid, &slb.vsid);
+ if (ret)
+ return ret;
spu_load_slb(spu, spu->slb_replace, &slb);
--
1.9.1
^ permalink raw reply related
* [PATCH 01/15] powerpc/cell: Move spu_handle_mm_fault() out of cell platform
From: Michael Neuling @ 2014-09-18 8:26 UTC (permalink / raw)
To: greg, arnd, mpe, benh
Cc: cbe-oss-dev, mikey, imunsie, linux-kernel, linuxppc-dev, jk,
anton
In-Reply-To: <1411028820-29933-1-git-send-email-mikey@neuling.org>
From: Ian Munsie <imunsie@au1.ibm.com>
Currently spu_handle_mm_fault() is in the cell platform.
This code is generically useful for other non-cell co-processors on powerpc.
This patch moves this function out of the cell platform into arch/powerpc/mm so
that others may use it.
Signed-off-by: Ian Munsie <imunsie@au1.ibm.com>
Signed-off-by: Michael Neuling <mikey@neuling.org>
---
arch/powerpc/include/asm/copro.h | 18 ++++++++++++++++++
arch/powerpc/include/asm/spu.h | 5 ++---
arch/powerpc/mm/Makefile | 1 +
.../{platforms/cell/spu_fault.c => mm/copro_fault.c} | 14 ++++++--------
arch/powerpc/platforms/cell/Makefile | 2 +-
arch/powerpc/platforms/cell/spufs/fault.c | 4 ++--
6 files changed, 30 insertions(+), 14 deletions(-)
create mode 100644 arch/powerpc/include/asm/copro.h
rename arch/powerpc/{platforms/cell/spu_fault.c => mm/copro_fault.c} (89%)
diff --git a/arch/powerpc/include/asm/copro.h b/arch/powerpc/include/asm/copro.h
new file mode 100644
index 0000000..2858108
--- /dev/null
+++ b/arch/powerpc/include/asm/copro.h
@@ -0,0 +1,18 @@
+/*
+ * Copyright 2014 IBM Corp.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ */
+
+#ifndef _ASM_POWERPC_COPRO_H
+#define _ASM_POWERPC_COPRO_H
+
+int copro_handle_mm_fault(struct mm_struct *mm, unsigned long ea,
+ unsigned long dsisr, unsigned *flt);
+
+int copro_data_segment(struct mm_struct *mm, u64 ea, u64 *esid, u64 *vsid);
+
+#endif /* _ASM_POWERPC_COPRO_H */
diff --git a/arch/powerpc/include/asm/spu.h b/arch/powerpc/include/asm/spu.h
index 37b7ca3..a6e6e2b 100644
--- a/arch/powerpc/include/asm/spu.h
+++ b/arch/powerpc/include/asm/spu.h
@@ -27,6 +27,8 @@
#include <linux/workqueue.h>
#include <linux/device.h>
#include <linux/mutex.h>
+#include <asm/reg.h>
+#include <asm/copro.h>
#define LS_SIZE (256 * 1024)
#define LS_ADDR_MASK (LS_SIZE - 1)
@@ -277,9 +279,6 @@ void spu_remove_dev_attr(struct device_attribute *attr);
int spu_add_dev_attr_group(struct attribute_group *attrs);
void spu_remove_dev_attr_group(struct attribute_group *attrs);
-int spu_handle_mm_fault(struct mm_struct *mm, unsigned long ea,
- unsigned long dsisr, unsigned *flt);
-
/*
* Notifier blocks:
*
diff --git a/arch/powerpc/mm/Makefile b/arch/powerpc/mm/Makefile
index d0130ff..a7f4dd7 100644
--- a/arch/powerpc/mm/Makefile
+++ b/arch/powerpc/mm/Makefile
@@ -34,3 +34,4 @@ obj-$(CONFIG_TRANSPARENT_HUGEPAGE) += hugepage-hash64.o
obj-$(CONFIG_PPC_SUBPAGE_PROT) += subpage-prot.o
obj-$(CONFIG_NOT_COHERENT_CACHE) += dma-noncoherent.o
obj-$(CONFIG_HIGHMEM) += highmem.o
+obj-$(CONFIG_SPU_BASE) += copro_fault.o
diff --git a/arch/powerpc/platforms/cell/spu_fault.c b/arch/powerpc/mm/copro_fault.c
similarity index 89%
rename from arch/powerpc/platforms/cell/spu_fault.c
rename to arch/powerpc/mm/copro_fault.c
index 641e727..ba7df14 100644
--- a/arch/powerpc/platforms/cell/spu_fault.c
+++ b/arch/powerpc/mm/copro_fault.c
@@ -1,5 +1,5 @@
/*
- * SPU mm fault handler
+ * CoProcessor (SPU/AFU) mm fault handler
*
* (C) Copyright IBM Deutschland Entwicklung GmbH 2007
*
@@ -23,16 +23,14 @@
#include <linux/sched.h>
#include <linux/mm.h>
#include <linux/export.h>
-
-#include <asm/spu.h>
-#include <asm/spu_csa.h>
+#include <asm/reg.h>
/*
* This ought to be kept in sync with the powerpc specific do_page_fault
* function. Currently, there are a few corner cases that we haven't had
* to handle fortunately.
*/
-int spu_handle_mm_fault(struct mm_struct *mm, unsigned long ea,
+int copro_handle_mm_fault(struct mm_struct *mm, unsigned long ea,
unsigned long dsisr, unsigned *flt)
{
struct vm_area_struct *vma;
@@ -58,12 +56,12 @@ int spu_handle_mm_fault(struct mm_struct *mm, unsigned long ea,
goto out_unlock;
}
- is_write = dsisr & MFC_DSISR_ACCESS_PUT;
+ is_write = dsisr & DSISR_ISSTORE;
if (is_write) {
if (!(vma->vm_flags & VM_WRITE))
goto out_unlock;
} else {
- if (dsisr & MFC_DSISR_ACCESS_DENIED)
+ if (dsisr & DSISR_PROTFAULT)
goto out_unlock;
if (!(vma->vm_flags & (VM_READ | VM_EXEC)))
goto out_unlock;
@@ -91,4 +89,4 @@ out_unlock:
up_read(&mm->mmap_sem);
return ret;
}
-EXPORT_SYMBOL_GPL(spu_handle_mm_fault);
+EXPORT_SYMBOL_GPL(copro_handle_mm_fault);
diff --git a/arch/powerpc/platforms/cell/Makefile b/arch/powerpc/platforms/cell/Makefile
index fe053e7..2d16884 100644
--- a/arch/powerpc/platforms/cell/Makefile
+++ b/arch/powerpc/platforms/cell/Makefile
@@ -20,7 +20,7 @@ spu-manage-$(CONFIG_PPC_CELL_COMMON) += spu_manage.o
obj-$(CONFIG_SPU_BASE) += spu_callbacks.o spu_base.o \
spu_notify.o \
- spu_syscalls.o spu_fault.o \
+ spu_syscalls.o \
$(spu-priv1-y) \
$(spu-manage-y) \
spufs/
diff --git a/arch/powerpc/platforms/cell/spufs/fault.c b/arch/powerpc/platforms/cell/spufs/fault.c
index 8cb6260..e45894a 100644
--- a/arch/powerpc/platforms/cell/spufs/fault.c
+++ b/arch/powerpc/platforms/cell/spufs/fault.c
@@ -138,7 +138,7 @@ int spufs_handle_class1(struct spu_context *ctx)
if (ctx->state == SPU_STATE_RUNNABLE)
ctx->spu->stats.hash_flt++;
- /* we must not hold the lock when entering spu_handle_mm_fault */
+ /* we must not hold the lock when entering copro_handle_mm_fault */
spu_release(ctx);
access = (_PAGE_PRESENT | _PAGE_USER);
@@ -149,7 +149,7 @@ int spufs_handle_class1(struct spu_context *ctx)
/* hashing failed, so try the actual fault handler */
if (ret)
- ret = spu_handle_mm_fault(current->mm, ea, dsisr, &flt);
+ ret = copro_handle_mm_fault(current->mm, ea, dsisr, &flt);
/*
* This is nasty: we need the state_mutex for all the bookkeeping even
--
1.9.1
^ permalink raw reply related
* [PATCH 0/15] POWER8 Coherent Accelerator device driver
From: Michael Neuling @ 2014-09-18 8:26 UTC (permalink / raw)
To: greg, arnd, mpe, benh
Cc: cbe-oss-dev, mikey, imunsie, linux-kernel, linuxppc-dev, jk,
anton
This add support for the Coherent Accelerator (cxl) attached to POWER8
processors. This coherent accelerator interface is designed to allow the
coherent connection of FPGA based accelerators (and other devices) to a POWER
systems.
IBM refers to this as the Coherent Accelerator Processor Interface or CAPI. In
this driver it's referred to by the name cxl to avoid confusion with the ISDN
CAPI subsystem.
An overview of the patches:
Patches 1-2: Split some of the old Cell co-processor code out so it can be
reused.
Patches 3-9: Add infrastructure to arch/powerpc needed by cxl.
Patches 10: Add call backs needed for invalidating cxl mm contexts.
Patch 11: Add cxl specific support that needs to be built in to the
kernel (can't be a module).
Patches 12-15: Add the majority of the device driver and API header.
Patch 15: Documentation.
The documentation in this last patch gives an overview of the hardware
architecture as well as the userspace API.
The cxl driver has a user-space interface described in include/uapi/misc/cxl.h
and Documentation/powerpc/cxl.txt. There are two ioctls which can be used to
talk to the driver once the new /dev/cxl/afu0.0 device is opened. This device
can also be read and mmaped.
There's also sysfs entries used to communicate information about the cxl
configuration to userspace. These are documented in
Documentation/ABI/testing/sysfs-class-cxl.
Many contributed to this device driver but Ian Munsie is the principal author.
Driver can also be found here (based on 3.17-rc5):
git://github.com/mikey/linux.git cxl
https://github.com/mikey/linux/commits/cxl
Please consider for inclusion. Feedback welcome!
Regards,
Mikey
Documentation/ABI/testing/sysfs-class-cxl | 125 ++++
Documentation/ioctl/ioctl-number.txt | 1 +
Documentation/powerpc/00-INDEX | 2 +
Documentation/powerpc/cxl.txt | 310 ++++++++
MAINTAINERS | 7 +
arch/powerpc/include/asm/copro.h | 18 +
arch/powerpc/include/asm/mmu-hash64.h | 3 +
arch/powerpc/include/asm/opal.h | 2 +
arch/powerpc/include/asm/pnv-pci.h | 27 +
arch/powerpc/include/asm/spu.h | 5 +-
arch/powerpc/mm/Makefile | 2 +
arch/powerpc/mm/copro_fault.c | 140 ++++
arch/powerpc/mm/hash_native_64.c | 6 +-
arch/powerpc/mm/hash_utils_64.c | 25 +-
arch/powerpc/mm/slb.c | 3 -
arch/powerpc/mm/slice.c | 3 +
arch/powerpc/platforms/cell/Makefile | 2 +-
arch/powerpc/platforms/cell/spu_base.c | 41 +-
arch/powerpc/platforms/cell/spu_fault.c | 94 ---
arch/powerpc/platforms/cell/spufs/fault.c | 4 +-
arch/powerpc/platforms/powernv/opal-wrappers.S | 1 +
arch/powerpc/platforms/powernv/pci-ioda.c | 229 +++++-
arch/powerpc/sysdev/msi_bitmap.c | 18 +-
drivers/misc/Kconfig | 1 +
drivers/misc/Makefile | 1 +
drivers/misc/cxl/Kconfig | 25 +
drivers/misc/cxl/Makefile | 4 +
drivers/misc/cxl/base.c | 102 +++
drivers/misc/cxl/context.c | 169 +++++
drivers/misc/cxl/cxl-pci.c | 977 +++++++++++++++++++++++++
drivers/misc/cxl/cxl.h | 605 +++++++++++++++
drivers/misc/cxl/debugfs.c | 116 +++
drivers/misc/cxl/fault.c | 298 ++++++++
drivers/misc/cxl/file.c | 503 +++++++++++++
drivers/misc/cxl/irq.c | 405 ++++++++++
drivers/misc/cxl/main.c | 238 ++++++
drivers/misc/cxl/native.c | 649 ++++++++++++++++
drivers/misc/cxl/sysfs.c | 348 +++++++++
include/misc/cxl.h | 34 +
include/uapi/Kbuild | 1 +
include/uapi/misc/Kbuild | 2 +
include/uapi/misc/cxl.h | 88 +++
42 files changed, 5463 insertions(+), 171 deletions(-)
^ permalink raw reply
* Re: [PATCH 0/9] powerpc/powernv: Support for fastsleep and winkle
From: Shreyas B Prabhu @ 2014-09-18 3:11 UTC (permalink / raw)
To: linux-kernel, Michael Ellerman
Cc: devicetree, Srivatsa S. Bhat, linux-pm, Rafael J. Wysocki,
Grant Likely, Rob Herring, Paul Mackerras, Preeti U. Murthy,
linuxppc-dev
In-Reply-To: <1408989673-14137-1-git-send-email-shreyas@linux.vnet.ibm.com>
Hi,
In this patch series we use winkle for offlined cores. I successfully
tested the working of this with subcore functionality.
Test scenario was as follows:
1. Set SMT mode to 1, Set subores-per-core to 1
2. Offline a core, in this case cpu 32 (sending it to winkle)
3. Set subcores-per-core to 4
4. Online the core
5. Start a guest (Topology 1 core 2 threads) on a subcore, in this case
on cpu 36
This works without any glitch.
Thanks,
Shreyas
On Monday 25 August 2014 11:31 PM, Shreyas B. Prabhu wrote:
> Fast sleep is an idle state, where the core and the L1 and L2
> caches are brought down to a threshold voltage. This also means that
> the communication between L2 and L3 caches have to be fenced. However
> the current P8 chips have a bug wherein this fencing between L2 and
> L3 caches get delayed by a cpu cycle. This can delay L3 response to
> the other cpus if they request for data during this time. Thus they
> would fetch the same data from the memory which could lead to data
> corruption if L3 cache is not flushed.
> Patch 4 adds support to work around this.
>
> 'Deep Winkle' is a deeper idle state where core and private L2 are powered
> off. While it offers higher power savings, it is at the cost of losing
> hypervisor register state and higher latency.
> Patch 5-9 adds support for winkle and uses it for offline cpus.
>
> Patch 1 - Moves parameters required discover idle states to a location
> common to both cpuidle driver and powernv core code
> Patch 2 - Populates idle state details from device tree
> Patch 3 - Enables cpus to run guest after waking up from fastsleep/winkle
>
>
> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
> Cc: Paul Mackerras <paulus@samba.org>
> Cc: Michael Ellerman <mpe@ellerman.id.au>
> Cc: Rafael J. Wysocki <rjw@rjwysocki.net>
> Cc: Srivatsa S. Bhat <srivatsa@MIT.EDU>
> Cc: Preeti U. Murthy <preeti@linux.vnet.ibm.com>
> Cc: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
> Cc: Rob Herring <robh+dt@kernel.org>
> Cc: Grant Likely <grant.likely@linaro.org>
> Cc: devicetree@vger.kernel.org
> Cc: linux-pm@vger.kernel.org
> Cc: linuxppc-dev@lists.ozlabs.org
>
> Preeti U Murthy (2):
> cpuidle/powernv: Populate cpuidle state details by querying the
> device-tree
> powerpc/powernv/cpuidle: Add workaround to enable fastsleep
>
> Shreyas B. Prabhu (6):
> powerpc/kvm/book3s_hv: Enable CPUs to run guest after waking up from
> fast-sleep
> powerpc/powernv: Add OPAL call to save and restore
> powerpc: Adding macro for accessing Thread Switch Control Register
> powerpc/powernv: Add winkle infrastructure
> powerpc/powernv: Discover and enable winkle
> powerpc/powernv: Enter deepest supported idle state in offline
>
> Srivatsa S. Bhat (1):
> powerpc/powernv: Enable Offline CPUs to enter deep idle states
>
> arch/powerpc/include/asm/machdep.h | 4 +
> arch/powerpc/include/asm/opal.h | 10 ++
> arch/powerpc/include/asm/paca.h | 3 +
> arch/powerpc/include/asm/ppc-opcode.h | 2 +
> arch/powerpc/include/asm/processor.h | 6 +-
> arch/powerpc/include/asm/reg.h | 1 +
> arch/powerpc/kernel/asm-offsets.c | 1 +
> arch/powerpc/kernel/exceptions-64s.S | 37 ++---
> arch/powerpc/kernel/idle.c | 30 ++++
> arch/powerpc/kernel/idle_power7.S | 83 +++++++++-
> arch/powerpc/platforms/powernv/opal-wrappers.S | 2 +
> arch/powerpc/platforms/powernv/powernv.h | 8 +
> arch/powerpc/platforms/powernv/setup.c | 217 +++++++++++++++++++++++++
> arch/powerpc/platforms/powernv/smp.c | 13 +-
> arch/powerpc/platforms/powernv/subcore.c | 15 ++
> drivers/cpuidle/cpuidle-powernv.c | 40 ++++-
> 16 files changed, 439 insertions(+), 33 deletions(-)
>
^ permalink raw reply
* Re: [PATCH v3 00/21] powerpc/8xx: Optimise MMU TLB handling and add support of 16k pages
From: Scott Wood @ 2014-09-17 20:34 UTC (permalink / raw)
To: christophe leroy; +Cc: linuxppc-dev, Paul Mackerras, linux-kernel
In-Reply-To: <5419F017.8010603@c-s.fr>
On Wed, 2014-09-17 at 22:33 +0200, christophe leroy wrote:
> Le 17/09/2014 18:40, Scott Wood a écrit :
> > On Wed, 2014-09-17 at 18:36 +0200, Christophe Leroy wrote:
> >> This patchset:
> >> 1) provides several MMU TLB handling optimisation on MPC8xx.
> >> 2) adds support of 16k pages on MPC8xx.
> >> All changes have been successfully tested on a custom board equipped with MPC885
> >>
> >> Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
> >> Tested-by: Christophe Leroy <christophe.leroy@c-s.fr>
> > I've already applied patches 1, 2, 4, 5, 6, 9, and 10 from the previous
> > patchset -- have they changed?
> >
> > -Scott
> >
> No, only 3, 7, 17 are changed, and 20,21 are new.
> I didn't notice you already applied some. How should I then proceed now
> for the remaining ones ? Submit a new set ?
>
No, I'll just skip the ones I've already applied.
-Scott
^ permalink raw reply
* Re: [PATCH v3 00/21] powerpc/8xx: Optimise MMU TLB handling and add support of 16k pages
From: christophe leroy @ 2014-09-17 20:33 UTC (permalink / raw)
To: Scott Wood; +Cc: linuxppc-dev, Paul Mackerras, linux-kernel
In-Reply-To: <1410972043.24184.532.camel@snotra.buserror.net>
Le 17/09/2014 18:40, Scott Wood a écrit :
> On Wed, 2014-09-17 at 18:36 +0200, Christophe Leroy wrote:
>> This patchset:
>> 1) provides several MMU TLB handling optimisation on MPC8xx.
>> 2) adds support of 16k pages on MPC8xx.
>> All changes have been successfully tested on a custom board equipped with MPC885
>>
>> Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
>> Tested-by: Christophe Leroy <christophe.leroy@c-s.fr>
> I've already applied patches 1, 2, 4, 5, 6, 9, and 10 from the previous
> patchset -- have they changed?
>
> -Scott
>
No, only 3, 7, 17 are changed, and 20,21 are new.
I didn't notice you already applied some. How should I then proceed now
for the remaining ones ? Submit a new set ?
Thanks
Christophe
---
Ce courrier électronique ne contient aucun virus ou logiciel malveillant parce que la protection avast! Antivirus est active.
http://www.avast.com
^ permalink raw reply
* Re: [5/5] pseries: Implement memory hotplug remove in the kernel
From: Nathan Fontenot @ 2014-09-17 19:58 UTC (permalink / raw)
To: Michael Ellerman; +Cc: linuxppc-dev list
In-Reply-To: <1410937650.27681.13.camel@concordia>
On 09/17/2014 02:07 AM, Michael Ellerman wrote:
>
> On Mon, 2014-09-15 at 15:33 -0500, Nathan Fontenot wrote:
>> This patch adds the ability to do memory hotplug remove in the kernel.
>>
>> Currently the hotplug add/remove of memory is handled by the drmgr
>> command. The drmgr command performs the add/remove by performing
>> some work in user-space and making requests to the kernel to handle
>> other pieces. By moving all of the work to the kernel we can do the
>> add and remove faster, and provide a common place to do memory hotplug
>> for both the PowerVM and PowerKVM environments.
>>
>> Signed-off-by: Nathan Fontenot <nfont@linux.vnet.ibm.com>
>> ---
>> arch/powerpc/platforms/pseries/hotplug-memory.c | 140 +++++++++++++++++++++++
>> 1 file changed, 139 insertions(+), 1 deletion(-)
>>
>> diff --git a/arch/powerpc/platforms/pseries/hotplug-memory.c b/arch/powerpc/platforms/pseries/hotplug-memory.c
>> index b254773..160c424 100644
>> --- a/arch/powerpc/platforms/pseries/hotplug-memory.c
>> +++ b/arch/powerpc/platforms/pseries/hotplug-memory.c
>> @@ -193,7 +193,137 @@ static int pseries_remove_mem_node(struct device_node *np)
>> pseries_remove_memblock(base, lmb_size);
>> return 0;
>> }
>> +
>> +static int lmb_is_removable(struct of_drconf_cell *lmb)
>> +{
>
> Do we not already have something like this?
No. Perhaps your thinking of the code in drivers/base/memory.c that
handles the sysfs removable file. That code just calls the same
is_mem_section_removable() routine.
>
>> + int i, scns_per_block;
>> + int rc = 1;
>
> I can see this makes the &= work below.
>
> But what if block_sz / MIN_MEMORY_BLOCK_SIZE = 0 ?
If that happens, something else is really wrong. Most
likely a malformed device tree.
For pseries MIN_MEMORY_BLOCK_SIZE is defined to be the smallest
LMB size we suppport, 16MB.
I can add a pr_warn() statement here and bail if that happens.
>
>> + unsigned long pfn, block_sz;
>> + u64 phys_addr;
>> +
>> + phys_addr = be64_to_cpu(lmb->base_addr);
>> + block_sz = memory_block_size_bytes();
>> + scns_per_block = block_sz / MIN_MEMORY_BLOCK_SIZE;
>> +
>> + for (i = 0; i < scns_per_block; i++) {
>> + pfn = PFN_DOWN(phys_addr);
>> + if (!pfn_present(pfn))
>> + continue;
>> +
>> + rc &= is_mem_section_removable(pfn, PAGES_PER_SECTION);
>> + phys_addr += MIN_MEMORY_BLOCK_SIZE;
>> + }
>> +
>> + return rc;
>> +}
>
>> +static int dlpar_memory_remove(struct pseries_hp_errorlog *hp_elog)
>> +{
>
> ...
>
>> +}
>
> Most of the same comments as for add.
>
ok, I'll go through them and apply them to the remove code.
Thanks for the review.
-Nathan
^ permalink raw reply
* Re: [RFC PATCH] dt:numa: adding numa node mapping for memory nodes.
From: Mark Rutland @ 2014-09-17 19:34 UTC (permalink / raw)
To: Kumar Gala
Cc: open list:OPEN FIRMWARE AND..., Pawel Moll, Ian Campbell,
Rob Herring, Ganapatrao Kulkarni, linuxppc-dev,
linux-arm-kernel@lists.infradead.org, gpkulkarni@gmail.com
In-Reply-To: <A4205FCA-D4DE-48A4-B19F-58400A3FBF5C@codeaurora.org>
On Wed, Sep 17, 2014 at 04:37:30PM +0100, Kumar Gala wrote:
>
> On Sep 17, 2014, at 1:56 AM, Ganapatrao Kulkarni <ganapatrao.kulkarni@caviumnetworks.com> wrote:
>
> > From: Ganapatrao Kulkarni <ganapatrao.kulkarni@cavium.com>
> >
> > This patch adds property "nid" to memory node to provide the memory range to
> > numa node id mapping.
> >
> > Signed-off-by: Ganapatrao Kulkarni <ganapatrao.kulkarni@cavium.com>
> >
> > —
>
> Adding the PPC guys as they’ve been doing NUMA on IBM Power Servers
> for years with OF/DT. So we should really try and follow what they’ve
> done.
Agreed.
> > Documentation/devicetree/bindings/numa.txt | 58 ++++++++++++++++++++++++++++++
> > 1 file changed, 58 insertions(+)
> > create mode 100644 Documentation/devicetree/bindings/numa.txt
> >
> > diff --git a/Documentation/devicetree/bindings/numa.txt b/Documentation/devicetree/bindings/numa.txt
> > new file mode 100644
> > index 0000000..c4a94f2
> > --- /dev/null
> > +++ b/Documentation/devicetree/bindings/numa.txt
> > @@ -0,0 +1,58 @@
> > +======================================================
> > +numa id binding description
> > +======================================================
> > +
> > +======================================================
> > +1 - Introduction
> > +======================================================
> > +The device node property "nid(numa node id)" can be added to memory
Why the quotes?
> > +device node to map the range of memory addresses as defined in property "reg".
> > +The property "nid" maps the memory range to the numa node id, which is used to
> > +find the local and remory pages on numa aware systems.
What is a "numa node id", exactly, and how is the OS intended to use it?
I don't see how this can be of any use as-is.
Mark.
> > +
> > +======================================================
> > +2 - nid property
> > +======================================================
> > +Numa node id, "nid" is required property of memory device node for
> > +numa enabled platforms.
> > +
> > +|------------------------------------------------------|
> > +|Property Type | Usage | Value Type | Definition |
> > +|------------------------------------------------------|
> > +| nid | R | <u32> | Numa Node id |
> > +| | | | for this memory |
> > +|------------------------------------------------------|
> > +
> > +========================================================
> > +4 - Example memory nodes with numa node id mapping
> > +========================================================
> > +
> > +Example 1 (2 memory nodes, each mapped to a numa node.):
> > +
> > + memory@00000000 {
> > + device_type = "memory";
> > + reg = <0x0 0x00000000 0x0 0x80000000>;
> > + nid = <0x0>;
> > + };
> > +
> > + memory@10000000000 {
> > + device_type = "memory";
> > + reg = <0x100 0x00000000 0x0 0x80000000>;
> > + nid = <0x1>;
> > + };
> > +
> > +Example 2 (multiple memory ranges in each memory node and mapped to numa node):
> > +
> > + memory@00000000 {
> > + device_type = "memory";
> > + reg = <0x0 0x00000000 0x0 0x80000000>,
> > + <0x0 0x80000000 0x0 0x80000000>;
> > + nid = <0x0>;
> > + };
> > +
> > + memory@10000000000 {
> > + device_type = "memory";
> > + reg = <0x100 0x00000000 0x0 0x80000000>,
> > + <0x100 0x80000000 0x0 0x80000000>;
> > + nid = <0x1>;
> > + };
> > --
> > 1.8.1.4
> >
>
> --
> Employee of Qualcomm Innovation Center, Inc.
> Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation
>
> --
> To unsubscribe from this list: send the line "unsubscribe devicetree" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
^ permalink raw reply
* Re: [4/5] pseries: Implement memory hotplug add in the kernel
From: Nathan Fontenot @ 2014-09-17 19:45 UTC (permalink / raw)
To: Michael Ellerman; +Cc: linuxppc-dev
In-Reply-To: <1410937639.27681.12.camel@concordia>
On 09/17/2014 02:07 AM, Michael Ellerman wrote:
>
> On Mon, 2014-09-15 at 15:32 -0500, Nathan Fontenot wrote:
>> This patch adds the ability to do memory hotplug adding in the kernel.
>>
>> Currently the hotplug add/remove of memory is handled by the drmgr
>> command. The drmgr command performs the add/remove by performing
>> some work in user-space and making requests to the kernel to handle
>> other pieces. By moving all of the work to the kernel we can do the
>> add and remove faster, and provide a common place to do memory hotplug
>> for both the PowerVM and PowerKVM environments.
>>
>> Signed-off-by: Nathan Fontenot <nfont@linux.vnet.ibm.com>
>> ---
>> arch/powerpc/platforms/pseries/hotplug-memory.c | 170 +++++++++++++++++++++++
>> 1 file changed, 170 insertions(+)
>>
>> diff --git a/arch/powerpc/platforms/pseries/hotplug-memory.c b/arch/powerpc/platforms/pseries/hotplug-memory.c
>> index 0e60e15..b254773 100644
>> --- a/arch/powerpc/platforms/pseries/hotplug-memory.c
>> +++ b/arch/powerpc/platforms/pseries/hotplug-memory.c
>> @@ -17,6 +17,7 @@
>> #include <linux/vmalloc.h>
>> #include <linux/memory.h>
>> #include <linux/memory_hotplug.h>
>> +#include <linux/slab.h>
>>
>> #include <asm/firmware.h>
>> #include <asm/machdep.h>
>> @@ -24,6 +25,8 @@
>> #include <asm/sparsemem.h>
>> #include <asm/rtas.h>
>>
>> +#include "pseries.h"
>> +
>> DEFINE_MUTEX(dlpar_mem_mutex);
>>
>> unsigned long pseries_memory_block_size(void)
>> @@ -69,6 +72,53 @@ unsigned long pseries_memory_block_size(void)
>> return memblock_size;
>> }
>>
>> +static void dlpar_free_drconf_property(struct property *prop)
>> +{
>> + kfree(prop->name);
>> + kfree(prop->value);
>> + kfree(prop);
>> +}
>> +
>> +static struct property *dlpar_clone_drconf_property(struct device_node *dn)
>> +{
>> + struct property *prop, *new_prop;
>> +
>> + prop = of_find_property(dn, "ibm,dynamic-memory", NULL);
>> + if (!prop)
>> + return NULL;
>> +
>> + new_prop = kzalloc(sizeof(*new_prop), GFP_KERNEL);
>> + if (!new_prop)
>> + return NULL;
>> +
>> + new_prop->name = kstrdup(prop->name, GFP_KERNEL);
>> + new_prop->value = kmalloc(prop->length + 1, GFP_KERNEL);
>> + if (!new_prop->name || !new_prop->value) {
>> + dlpar_free_drconf_property(new_prop);
>> + return NULL;
>> + }
>> +
>> + memcpy(new_prop->value, prop->value, prop->length);
>> + new_prop->length = prop->length;
>> + *(((char *)new_prop->value) + new_prop->length) = 0;
>
> It's not a string, is it?
No, property->value is a void*. I'll drop that line of code.
>
>> + return new_prop;
>> +}
>> +
>> +static struct memory_block *lmb_to_memblock(struct of_drconf_cell *lmb)
>> +{
>> + unsigned long section_nr;
>> + struct mem_section *mem_sect;
>> + struct memory_block *mem_block;
>> + u64 phys_addr = be64_to_cpu(lmb->base_addr);
>> +
>> + section_nr = pfn_to_section_nr(PFN_DOWN(phys_addr));
>> + mem_sect = __nr_to_section(section_nr);
>> +
>> + mem_block = find_memory_block(mem_sect);
>> + return mem_block;
>> +}
>> +
>> #ifdef CONFIG_MEMORY_HOTREMOVE
>> static int pseries_remove_memory(u64 start, u64 size)
>> {
>> @@ -155,13 +205,133 @@ static inline int pseries_remove_mem_node(struct device_node *np)
>> }
>> #endif /* CONFIG_MEMORY_HOTREMOVE */
>>
>> +static int dlpar_add_one_lmb(struct of_drconf_cell *lmb)
>> +{
>> + struct memory_block *mem_block;
>> + u64 phys_addr;
>> + unsigned long pages_per_block;
>> + unsigned long block_sz;
>> + int nid, sections_per_block;
>> + int rc;
>> +
>> + phys_addr = be64_to_cpu(lmb->base_addr);
>
> of_drconf_cell needs endian annotations.
Yes it does. I can include a patch to update the struct.
>
>> + block_sz = memory_block_size_bytes();
>> + sections_per_block = block_sz / MIN_MEMORY_BLOCK_SIZE;
>> + pages_per_block = PAGES_PER_SECTION * sections_per_block;
>> +
>> + if (phys_addr & ((pages_per_block << PAGE_SHIFT) - 1))
>> + return -EINVAL;
>> +
>> + nid = memory_add_physaddr_to_nid(phys_addr);
>> + rc = add_memory(nid, phys_addr, block_sz);
>> + if (rc)
>> + return rc;
>> +
>> + rc = memblock_add(phys_addr, block_sz);
>> + if (rc) {
>> + remove_memory(nid, phys_addr, block_sz);
>> + return rc;
>> + }
>> +
>> + mem_block = lmb_to_memblock(lmb);
>> + if (!mem_block) {
>> + remove_memory(nid, phys_addr, block_sz);
>> + return -EINVAL;
>> + }
>
> That could all use a lot of comments. ie. why do we have to add it twice?
We don't actually add it twice, though I can see how one could think
that based on the names of the routines called. I'll add comments to
clarify this in v2 of the patch.
memory_add_physaddr_to_nid(), this doesn't add anything despite its naming.
The routine finds the node id for the specified physical address.
add_memory(), this actually adds the memory.
memblock_add(), this informs the memory block information tracking about the
newly added memory. Why this is not done as part of add_memory I don't know.
>
>> + rc = device_online(&mem_block->dev);
>> + put_device(&mem_block->dev);
>> + if (rc)
>> + remove_memory(nid, phys_addr, block_sz);
>> +
>> + return rc;
>> +}
>> +
>> +static int dlpar_memory_add(struct pseries_hp_errorlog *hp_elog)
>> +{
>> + struct of_drconf_cell *lmb;
>> + struct device_node *dn;
>> + struct property *prop;
>> + uint32_t entries, *p;
>
> *p should be __be32.
Yes.
>
>> + int i, lmbs_to_add;
>> + int lmbs_added = 0;
>> + int rc = -EINVAL;
>
> Don't pre-initialise your rc variables.
I did this here for a reason. When asking to add memory by drc_index it
is possible to fall out of the for() loop traversing the lmb entries
and not find the requested drc_index.
Adding a check for this situation after the loop would do the same thing
and probably make this situation more clear.
>
>> + if (hp_elog->id_type == PSERIES_HP_ELOG_ID_DRC_COUNT) {
>> + lmbs_to_add = be32_to_cpu(hp_elog->_drc_u.drc_count);
>> + pr_info("Attempting to hot-add %d LMB(s)\n", lmbs_to_add);
>> + } else {
>> + lmbs_to_add = 1;
>> + pr_info("Attempting to hot-add LMB, drc index %x\n",
>> + be32_to_cpu(hp_elog->_drc_u.drc_index));
>> + }
>> +
>> + dn = of_find_node_by_path("/ibm,dynamic-reconfiguration-memory");
>> + if (!dn)
>> + return -EINVAL;
>> +
>> + prop = dlpar_clone_drconf_property(dn);
>> + if (!prop) {
>> + of_node_put(dn);
>> + return -EINVAL;
>> + }
>> +
>> + p = prop->value;
>> + entries = be32_to_cpu(*p++);
>> + lmb = (struct of_drconf_cell *)p;
>
>
> So if I'm reading this right the hp_elog either contains an index or a count of
> LMBs to add. But it doesn't contain anything about which address ranges to add
> or any of those details. That is all in the ibm,dynamic-memory property - but
> how did it get in there?
The ibm,dynamic-memory property of the device tree is passed to the kernel
by phyp/qemu. The property is an array that associates each LMB with a
starting physical address and associativity.
>
>> +
>> + for (i = 0; i < entries; i++, lmb++) {
>> + u32 drc_index = be32_to_cpu(lmb->drc_index);
>> +
>> + if (lmbs_to_add == lmbs_added)
>> + break;
>> +
>> + if (be32_to_cpu(lmb->flags) & DRCONF_MEM_ASSIGNED)
>> + continue;
>> +
>> + if (hp_elog->id_type == PSERIES_HP_ELOG_ID_DRC_INDEX
>> + && lmb->drc_index != hp_elog->_drc_u.drc_index)
>> + continue;
>> +
>> + rc = dlpar_acquire_drc(drc_index);
>> + if (rc)
>> + continue;
>> +
>> + rc = dlpar_add_one_lmb(lmb);
>> + if (rc) {
>> + dlpar_release_drc(drc_index);
>> + continue;
>> + }
>
> In both the above error cases you just move along. That means we potentially
> hotplugged some memory but not everything that we were asked to. That seems
> like a bad idea, we should either do everything or nothing.
>
>
That can be done, though will require some additional tracking.
The current hotplug handling for PowerVM make a best effort and tries to
add/remove as much of the requested memory as possible. I was going with
that same approach here, but have no problem moving to an all or nothing
approach.
We will need to keep track of the LMBs added/removed during a request so
we can return to the original state if the request cannot be satisfied.
>> +
>> + lmb->flags |= cpu_to_be32(DRCONF_MEM_ASSIGNED);
>> + lmbs_added++;
>> + pr_info("Memory at %llx (drc index %x) has been hot-added\n",
>> + be64_to_cpu(lmb->base_addr), drc_index);
>> + }
>> +
>> + if (lmbs_added)
>> + rc = of_update_property(dn, prop);
>> + else
>> + dlpar_free_drconf_property(prop);
>
> The value of rc here is not clear. It could be EINVAL or it could be the result
> of the last dlpar_add_one_lmb(lmb). gcc would have told you that if you hadn't
> initialised it.
>
>> +
>> + of_node_put(dn);
>> + return rc ? rc : lmbs_added;
>
> This looks wrong.
>
> Doesn't the rc eventually go back to dlpar_write(), which expects 0 for success?
>
> That should show up as the write failing in userspace.
>
>
Based on previous comments I think the handling of rc will be updated so
we either return success or failure.
>> int dlpar_memory(struct pseries_hp_errorlog *hp_elog)
>> {
>> int rc = 0;
>
> Don't initialise to zero, that way gcc can tell you if there's a path where you
> forget to initialise it. It also means you can't accidentally return success.
>
>> + if (hp_elog->id_type != PSERIES_HP_ELOG_ID_DRC_COUNT
>> + && hp_elog->id_type != PSERIES_HP_ELOG_ID_DRC_INDEX)
>> + return -EINVAL;
>
> This would look nicer as a switch I think.
can do.
-Nathan
^ permalink raw reply
* Re: [3/5] pseries: Create device hotplug entry point
From: Nathan Fontenot @ 2014-09-17 19:15 UTC (permalink / raw)
To: Michael Ellerman; +Cc: linuxppc-dev
In-Reply-To: <1410937632.27681.11.camel@concordia>
On 09/17/2014 02:07 AM, Michael Ellerman wrote:
>
> On Mon, 2014-09-15 at 15:31 -0500, Nathan Fontenot wrote:
>> For pseries system the kernel will be notified of hotplug requests in
>> the form of rtas hotplug events.
>
> Can you flesh that design out a bit for me, I don't entirely get how it's going
> to work.
>
> The kernel gets the rtas hotplug events (in rtasd.c) and spits them out to
> userspace, which then writes them back in ?
>
>> This patch creates a common routine that can handle these requests in both
>> the PowerVM anbd PowerKVM environments, handle_dlpar_errorlog(). This also
> ^
>> creates the initial memory hotplug request handling stub.
>>
>> For PowerVM this patch also creates a new /proc file that the drmgr
>> command will use to write rtas hotplug events to.
>
> Why is this different between phyp and KVM?
>
>> For future PowerKVM handling the rtas check-exception code can pass
>> any rtas hotplug events received to handle_dlpar_errorlog().
>
> Internally to the kernel you mean?
>
Perhaps a better explanation of how things work today and where I see
them going is needed. I was trying to avoid a long explanation and I
don't think my shortened explanation worked. I'll include this in v2
of the patchset too.
The current hotplug (or dlpar) of devices (the process is generally the
same for memory, cpu, and pci) on PowerVM systems is initiated
from the HMC, which communicates the request to the partitions through
the RSCT framework. The RSCT framework then invokes the drmgr command.
The drmgr command performs the hotplug operation by doing some pieces,
such as most of the rtas calls and device tree parsing, in userspace
and make requests to the kernel to online/offline the device, update the
device tree and add/remove the device.
For PowerKVM the approach is to follow what is currently being done for
pci hotplug. A hotplug request is initiated from the host. QEMU then
sends an EPOW interrupt to the guest which causes the guest to make the
rtas,check-exception call. In QEMU, the rtas,check-exception call
returns a rtas hotplug event to the guest. I was using this same framework
to also enable memory (and next cpu) hotplug.
You are correct that the current pci hotplug path for PowerKVM involves
the kernel receiving the rtas event, passing it to rtas_errd in userspace,
and having rtas_errd invoke drmgr. The drmgr command then handles the request
as described above for PowerVM systems.
There is no need for this circuitous route, we should just handle the entire
hotplug of devices in the kernel. What I am hoping to do is to enable this
by moving the code to handle hotplug from drmgr into the kernel and
provide a single path for handling hotplug for PowerVM and PowerKVM. To
make this work for PowerKVM we will update the kernel rtas code to
recognize rtas hotplug events returned from rtas,check-exception calls
and call handle_dlpar_errorlog(). The hotplug rtas event is never sent out
to userspace.
For PowerVM systems, I created the /proc/powerpc/dlpar file that a rtas
hotplug event can be written to and passed to handle_dlpar_errorlog().
There is no chance of updating how we receive hotplug requests on PowerVM
systems.
Hopefully that explains the design better.
>> diff --git a/arch/powerpc/platforms/pseries/dlpar.c b/arch/powerpc/platforms/pseries/dlpar.c
>> index a2450b8..574ec73 100644
>> --- a/arch/powerpc/platforms/pseries/dlpar.c
>> +++ b/arch/powerpc/platforms/pseries/dlpar.c
>> @@ -16,7 +16,9 @@
>> #include <linux/cpu.h>
>> #include <linux/slab.h>
>> #include <linux/of.h>
>> +#include <linux/proc_fs.h>
>> #include "offline_states.h"
>> +#include "pseries.h"
>>
>> #include <asm/prom.h>
>> #include <asm/machdep.h>
>> @@ -530,13 +532,72 @@ static ssize_t dlpar_cpu_release(const char *buf, size_t count)
>> return count;
>> }
>>
>> +#endif /* CONFIG_ARCH_CPU_PROBE_RELEASE */
>
> That is really confusing, but I think it's just a diff artifact?
Yes, diff artifact.
>
>> +static int handle_dlpar_errorlog(struct rtas_error_log *error_log)
>> +{
>> + struct pseries_errorlog *pseries_log;
>> + struct pseries_hp_errorlog *hp_elog;
>> + int rc = -EINVAL;
>> +
>> + pseries_log = get_pseries_errorlog(error_log,
>> + PSERIES_ELOG_SECT_ID_HOTPLUG);
>> + if (!pseries_log)
>> + return rc;
>> +
>> + hp_elog = (struct pseries_hp_errorlog *)pseries_log->data;
>> + if (!hp_elog)
>> + return rc;
>
> I don't see how that can happen?
>
> struct pseries_errorlog {
> __be16 id; /* 0x00 2-byte ASCII section ID */
> __be16 length; /* 0x02 Section length in bytes */
> uint8_t version; /* 0x04 Section version */
> uint8_t subtype; /* 0x05 Section subtype */
> __be16 creator_component; /* 0x06 Creator component ID */
> uint8_t data[]; /* 0x08 Start of section data */
> };
>
> Should you be checking for length == 0 instead ?
>
You are correct.
> Also I think the code will probably end up cleaner if you do the endian
> conversions immediately when you read the hp_elog, rather than passing it
> around in BE and having to remember to convert at all the usages.
>
Agreed. I'll clean that up.
>> + switch (hp_elog->resource) {
>> + case PSERIES_HP_ELOG_RESOURCE_MEM:
>> + rc = dlpar_memory(hp_elog);
>> + break;
>
> Please add:
>
> default:
> pr_warn_ratelimited("Unknown resource ..")
>
> Or something.
can do.
>
>
>> + }
>> +
>> + return rc;
>> +}
>> +
>> +static ssize_t dlpar_write(struct file *file, const char __user *buf,
>> + size_t count, loff_t *offset)
>> +{
>> + char *event_buf;
>> + int rc;
>> +
>> + event_buf = kmalloc(count + 1, GFP_KERNEL);
>
> Why + 1 ? It's not null-terminated AFAICS.
I think that's just habit. You're correct, it's not needed.
>
>> + if (!event_buf)
>> + return -ENOMEM;
>> +
>> + rc = copy_from_user(event_buf, buf, count);
>> + if (rc) {
>> + kfree(event_buf);
>> + return rc;
>> + }
>> +
>> + rc = handle_dlpar_errorlog((struct rtas_error_log *)event_buf);
>
> If you start with a struct rtas_error_log * you shouldn't need any casts.
good point.
>
>> + kfree(event_buf);
>> + return rc ? rc : count;
>> +}
>> +
>> +static const struct file_operations dlpar_fops = {
>> + .write = dlpar_write,
>> + .llseek = noop_llseek,
>> +};
>> +
>> static int __init pseries_dlpar_init(void)
>> {
>> + struct proc_dir_entry *proc_ent;
>> +
>> + proc_ent = proc_create("powerpc/dlpar", S_IWUSR, NULL, &dlpar_fops);
>> + if (proc_ent)
>> + proc_set_size(proc_ent, 0);
>
> else
> error message at least please
>
> Why are we putting it in /proc, can't it go in /sys/kernel like the mobility
> stuff?
I have no personal preference, I'll move it to /sys/kernel and update the
creation handling.
>
>> diff --git a/arch/powerpc/platforms/pseries/hotplug-memory.c b/arch/powerpc/platforms/pseries/hotplug-memory.c
>> index 24abc5c..0e60e15 100644
>> --- a/arch/powerpc/platforms/pseries/hotplug-memory.c
>> +++ b/arch/powerpc/platforms/pseries/hotplug-memory.c
>> @@ -20,6 +22,9 @@
>> #include <asm/machdep.h>
>> #include <asm/prom.h>
>> #include <asm/sparsemem.h>
>> +#include <asm/rtas.h>
>> +
>> +DEFINE_MUTEX(dlpar_mem_mutex);
>
> static ?
yes, that should have been static.
>
>> diff --git a/arch/powerpc/platforms/pseries/pseries.h b/arch/powerpc/platforms/pseries/pseries.h
>> index b94516b..28bd994 100644
>> --- a/arch/powerpc/platforms/pseries/pseries.h
>> +++ b/arch/powerpc/platforms/pseries/pseries.h
>> @@ -62,6 +63,15 @@ extern int dlpar_detach_node(struct device_node *);
>> extern int dlpar_acquire_drc(u32);
>> extern int dlpar_release_drc(u32);
>>
>> +#ifdef CONFIG_MEMORY_HOTPLUG
>> +extern int dlpar_memory(struct pseries_hp_errorlog *);
>> +#else
>> +static inline int dlpar_memory(struct pseries_hp_errorlog *hp_elog)
>> +{
>> + return -ENOTSUPP;
>
> EOPNOTSUPP is a bit more standard.
ok.
Thanks for all the feedback.
-Nathan
^ permalink raw reply
* Re: [PATCH v4 04/11] drivers: base: support cpu cache information interface to userspace via sysfs
From: Greg Kroah-Hartman @ 2014-09-17 19:00 UTC (permalink / raw)
To: Sudeep Holla
Cc: linux-s390@vger.kernel.org, Lorenzo Pieralisi,
linux-ia64@vger.kernel.org, Heiko Carstens, x86@kernel.org,
Stephen Boyd, LKML, linux-api@vger.kernel.org,
linux390@de.ibm.com, linuxppc-dev@lists.ozlabs.org,
linux-arm-kernel@lists.infradead.org
In-Reply-To: <5419C3F6.4050607@arm.com>
On Wed, Sep 17, 2014 at 06:25:10PM +0100, Sudeep Holla wrote:
> Hi Greg,
>
> On 03/09/14 18:00, Sudeep Holla wrote:
> >From: Sudeep Holla <sudeep.holla@arm.com>
> >
> >This patch adds initial support for providing processor cache information
> >to userspace through sysfs interface. This is based on already existing
> >implementations(x86, ia64, s390 and powerpc) and hence the interface is
> >intended to be fully compatible.
> >
> >The main purpose of this generic support is to avoid further code
> >duplication to support new architectures and also to unify all the existing
> >different implementations.
> >
> >This implementation maintains the hierarchy of cache objects which reflects
> >the system's cache topology. Cache devices are instantiated as needed as
> >CPUs come online. The cache information is replicated per-cpu even if they are
> >shared. A per-cpu array of cache information maintained is used mainly for
> >sysfs-related book keeping.
> >
> >It also implements the shared_cpu_map attribute, which is essential for
> >enabling both kernel and user-space to discover the system's overall cache
> >topology.
> >
> >This patch also add the missing ABI documentation for the cacheinfo sysfs
> >interface already, which is well defined and widely used.
> >
>
> Can you review the first 4 patches in this series please ?
It's in my todo queue, which is really long at the moment due to me
going to conferences (at one right now...) Will be working on this
soon, thanks for your patience.
greg k-h
^ permalink raw reply
* [PATCH v3 10/10] powerpc/perf/hv-24x7: Document sysfs event description entries
From: Sukadev Bhattiprolu @ 2014-09-17 18:52 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, ak, Jiri Olsa, peterz, eranian,
Paul Mackerras
Cc: Michael Ellerman, linux-kernel, dev, linuxppc-dev,
Anshuman Khandual
In-Reply-To: <1410979923-7722-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
.../testing/sysfs-bus-event_source-devices-hv_24x7 | 22 ++++++++++++++++++++
1 file changed, 22 insertions(+)
diff --git a/Documentation/ABI/testing/sysfs-bus-event_source-devices-hv_24x7 b/Documentation/ABI/testing/sysfs-bus-event_source-devices-hv_24x7
index e78ee79..5b501d7 100644
--- a/Documentation/ABI/testing/sysfs-bus-event_source-devices-hv_24x7
+++ b/Documentation/ABI/testing/sysfs-bus-event_source-devices-hv_24x7
@@ -21,3 +21,25 @@ Contact: Cody P Schafer <cody@linux.vnet.ibm.com>
Description:
Exposes the "version" field of the 24x7 catalog. This is also
extractable from the provided binary "catalog" sysfs entry.
+
+What: /sys/bus/event_source/devices/hv_24x7/event_descs/<event-name>
+Date: February 2014
+Contact: Cody P Schafer <cody@linux.vnet.ibm.com>
+Description:
+ Provides the description of a particular event as provided by
+ the firmware. If firmware does not provide a description, no
+ file will be created.
+
+ Note that the event-name lacks the domain suffix appended for
+ events in the events/ dir.
+
+What: /sys/bus/event_source/devices/hv_24x7/event_long_descs/<event-name>
+Date: February 2014
+Contact: Cody P Schafer <cody@linux.vnet.ibm.com>
+Description:
+ Provides the "long" description of a particular event as
+ provided by the firmware. If firmware does not provide a
+ description, no file will be created.
+
+ Note that the event-name lacks the domain suffix appended for
+ events in the events/ dir.
--
1.7.9.5
^ permalink raw reply related
* [PATCH v3 09/10] tools/perf: Document parameterized and symbolic events
From: Sukadev Bhattiprolu @ 2014-09-17 18:52 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, ak, Jiri Olsa, peterz, eranian,
Paul Mackerras
Cc: Michael Ellerman, linux-kernel, dev, linuxppc-dev,
Anshuman Khandual
In-Reply-To: <1410979923-7722-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
Changelog[v6]:
- [Sukadev Bhattiprolu]: Update documentation of perf-list and
perf-record; Added documentation for perf-stat.
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
---
tools/perf/Documentation/perf-list.txt | 13 +++++++++++++
tools/perf/Documentation/perf-record.txt | 12 ++++++++++++
tools/perf/Documentation/perf-stat.txt | 20 ++++++++++++++++----
3 files changed, 41 insertions(+), 4 deletions(-)
diff --git a/tools/perf/Documentation/perf-list.txt b/tools/perf/Documentation/perf-list.txt
index 6fce6a6..c405da24 100644
--- a/tools/perf/Documentation/perf-list.txt
+++ b/tools/perf/Documentation/perf-list.txt
@@ -89,6 +89,19 @@ raw encoding of 0x1A8 can be used:
You should refer to the processor specific documentation for getting these
details. Some of them are referenced in the SEE ALSO section below.
+PARAMETERIZED EVENTS
+--------------------
+
+Some pmu events listed by 'perf-list' will be displayed with '?' in them. For
+example:
+
+ hv_gpci/dtbp_ptitc,starting_index=?/
+
+This means that when provided as an event, a value for 'starting_index' must
+also be supplied. For example:
+
+ perf stat -C 0 -e 'hv_gpci/dtbp_ptitc,starting_index=0x2/' ...
+
OPTIONS
-------
diff --git a/tools/perf/Documentation/perf-record.txt b/tools/perf/Documentation/perf-record.txt
index d460049..a6a2f9c 100644
--- a/tools/perf/Documentation/perf-record.txt
+++ b/tools/perf/Documentation/perf-record.txt
@@ -33,6 +33,18 @@ OPTIONS
- a raw PMU event (eventsel+umask) in the form of rNNN where NNN is a
hexadecimal event descriptor.
+ - a symbolically formed PMU event like 'pmu/param1=0x3,param2/' where
+ 'param1', 'param2', etc are defined as formats for the PMU in
+ /sys/bus/event_sources/devices/<pmu>/format/*.
+
+ - a symbolically formed event like 'pmu/config=M,config1=N,config3=K/'
+
+ where M, N, K are numbers (in decimal, hex, octal format). Acceptable
+ values for each of 'config', 'config1' and 'config2' are defined by
+ corresponding entries in /sys/bus/event_sources/devices/<pmu>/format/*
+ param1 and param2 are defined as formats for the PMU in:
+ /sys/bus/event_sources/devices/<pmu>/format/*
+
- a hardware breakpoint event in the form of '\mem:addr[:access]'
where addr is the address in memory you want to break in.
Access is the memory access type (read, write, execute) it can
diff --git a/tools/perf/Documentation/perf-stat.txt b/tools/perf/Documentation/perf-stat.txt
index 29ee857..04e150d 100644
--- a/tools/perf/Documentation/perf-stat.txt
+++ b/tools/perf/Documentation/perf-stat.txt
@@ -25,10 +25,22 @@ OPTIONS
-e::
--event=::
- Select the PMU event. Selection can be a symbolic event name
- (use 'perf list' to list all events) or a raw PMU
- event (eventsel+umask) in the form of rNNN where NNN is a
- hexadecimal event descriptor.
+ Select the PMU event. Selection can be:
+
+ - a symbolic event name (use 'perf list' to list all events)
+
+ - a raw PMU event (eventsel+umask) in the form of rNNN where NNN is a
+ hexadecimal event descriptor.
+
+ - a symbolically formed event like 'pmu/param1=0x3,param2/' where
+ param1 and param2 are defined as formats for the PMU in
+ /sys/bus/event_sources/devices/<pmu>/format/*
+
+ - a symbolically formed event like 'pmu/config=M,config1=N,config2=K/'
+ where M, N, K are numbers (in decimal, hex, octal format).
+ Acceptable values for each of 'config', 'config1' and 'config2'
+ parameters are defined by corresponding entries in
+ /sys/bus/event_sources/devices/<pmu>/format/*
-i::
--no-inherit::
--
1.7.9.5
^ permalink raw reply related
* [PATCH v3 08/10] perf Documentation: add event parameters
From: Sukadev Bhattiprolu @ 2014-09-17 18:52 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, ak, Jiri Olsa, peterz, eranian,
Paul Mackerras
Cc: Michael Ellerman, linux-kernel, dev, linuxppc-dev,
Anshuman Khandual
In-Reply-To: <1410979923-7722-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
Event parameters are a basic way for partial events to be specified in
sysfs with per-event names given to the fields that need to be filled in
when using a particular event.
It is intended for supporting cases where the single 'cpu' parameter is
insufficient. For example, POWER 8 has events for physical
sockets/cores/cpus that are accessible from with virtual machines. To
keep using the single 'cpu' parameter we'd need to perform a mapping
between Linux's cpus and the physical machine's cpus (in this case
Linux is running under a hypervisor). This isn't possible because
bindings between our cpus and physical cpus may not be fixed, and we
probably won't have a "cpu" on each physical cpu.
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
.../testing/sysfs-bus-event_source-devices-events | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/Documentation/ABI/testing/sysfs-bus-event_source-devices-events b/Documentation/ABI/testing/sysfs-bus-event_source-devices-events
index 20979f8..c1f9850 100644
--- a/Documentation/ABI/testing/sysfs-bus-event_source-devices-events
+++ b/Documentation/ABI/testing/sysfs-bus-event_source-devices-events
@@ -52,12 +52,18 @@ Description: Per-pmu performance monitoring events specific to the running syste
event=0x2abc
event=0x423,inv,cmask=0x3
domain=0x1,offset=0x8,starting_index=0xffff
+ domain=0x1,offset=0x8,starting_index=phys_cpu
Each of the assignments indicates a value to be assigned to a
particular set of bits (as defined by the format file
corresponding to the <term>) in the perf_event structure passed
to the perf_open syscall.
+ In the case of the last example, a value replacing "phys_cpu"
+ would need to be provided by the user selecting the particular
+ event. This is refered to as "event parameterization". All
+ non-numerical values indicate an event parameter.
+
What: /sys/bus/event_source/devices/<pmu>/events/<event>.unit
Date: 2014/02/24
Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
--
1.7.9.5
^ permalink raw reply related
* [PATCH v3 06/10] powerpc/perf/{hv-gpci, hv-common}: generate requests with counters annotated
From: Sukadev Bhattiprolu @ 2014-09-17 18:51 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, ak, Jiri Olsa, peterz, eranian,
Paul Mackerras
Cc: Michael Ellerman, linux-kernel, dev, linuxppc-dev,
Anshuman Khandual
In-Reply-To: <1410979923-7722-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
This adds (in req-gen/) a framework for defining gpci counter requests.
It uses macro magic similar to ftrace.
Also convert the existing hv-gpci request structures and enum values to
use the new framework (and adjust old users of the structs and enum
values to cope with changes in naming).
In exchange for this macro disaster, we get autogenerated event listing
for GPCI in sysfs, build time field offset checking, and zero
duplication of information about GPCI requests.
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
arch/powerpc/perf/hv-common.c | 10 +-
arch/powerpc/perf/hv-gpci-requests.h | 79 ++++++++++++++
arch/powerpc/perf/hv-gpci.c | 8 ++
arch/powerpc/perf/hv-gpci.h | 37 +++----
arch/powerpc/perf/req-gen/_begin.h | 13 +++
arch/powerpc/perf/req-gen/_clear.h | 5 +
arch/powerpc/perf/req-gen/_end.h | 4 +
arch/powerpc/perf/req-gen/_request-begin.h | 15 +++
arch/powerpc/perf/req-gen/_request-end.h | 8 ++
arch/powerpc/perf/req-gen/perf.h | 155 ++++++++++++++++++++++++++++
10 files changed, 304 insertions(+), 30 deletions(-)
create mode 100644 arch/powerpc/perf/hv-gpci-requests.h
create mode 100644 arch/powerpc/perf/req-gen/_begin.h
create mode 100644 arch/powerpc/perf/req-gen/_clear.h
create mode 100644 arch/powerpc/perf/req-gen/_end.h
create mode 100644 arch/powerpc/perf/req-gen/_request-begin.h
create mode 100644 arch/powerpc/perf/req-gen/_request-end.h
create mode 100644 arch/powerpc/perf/req-gen/perf.h
diff --git a/arch/powerpc/perf/hv-common.c b/arch/powerpc/perf/hv-common.c
index 47e02b3..7dce8f10 100644
--- a/arch/powerpc/perf/hv-common.c
+++ b/arch/powerpc/perf/hv-common.c
@@ -9,13 +9,13 @@ unsigned long hv_perf_caps_get(struct hv_perf_caps *caps)
unsigned long r;
struct p {
struct hv_get_perf_counter_info_params params;
- struct cv_system_performance_capabilities caps;
+ struct hv_gpci_system_performance_capabilities caps;
} __packed __aligned(sizeof(uint64_t));
struct p arg = {
.params = {
.counter_request = cpu_to_be32(
- CIR_SYSTEM_PERFORMANCE_CAPABILITIES),
+ HV_GPCI_system_performance_capabilities),
.starting_index = cpu_to_be32(-1),
.counter_info_version_in = 0,
}
@@ -31,9 +31,9 @@ unsigned long hv_perf_caps_get(struct hv_perf_caps *caps)
caps->version = arg.params.counter_info_version_out;
caps->collect_privileged = !!arg.caps.perf_collect_privileged;
- caps->ga = !!(arg.caps.capability_mask & CV_CM_GA);
- caps->expanded = !!(arg.caps.capability_mask & CV_CM_EXPANDED);
- caps->lab = !!(arg.caps.capability_mask & CV_CM_LAB);
+ caps->ga = !!(arg.caps.capability_mask & HV_GPCI_CM_GA);
+ caps->expanded = !!(arg.caps.capability_mask & HV_GPCI_CM_EXPANDED);
+ caps->lab = !!(arg.caps.capability_mask & HV_GPCI_CM_LAB);
return r;
}
diff --git a/arch/powerpc/perf/hv-gpci-requests.h b/arch/powerpc/perf/hv-gpci-requests.h
new file mode 100644
index 0000000..0dfc4d9
--- /dev/null
+++ b/arch/powerpc/perf/hv-gpci-requests.h
@@ -0,0 +1,79 @@
+
+#include "req-gen/_begin.h"
+
+/*
+ * Based on the document "getPerfCountInfo v1.07"
+ */
+
+/* this needs to be -1 encoded in hex suitable for parsing by tools/perf. */
+#define M1 0xffffffffffffffff
+
+/*
+ * #define REQUEST_NAME counter_request_name
+ * #define REQUEST_NUM r_num
+ * #define REQUEST_IDX_KIND starting_index_kind
+ * #include I(REQUEST_BEGIN)
+ * REQUEST(
+ * __field(...)
+ * __field(...)
+ * __array(...)
+ * __count(...)
+ * )
+ * #include I(REQUEST_END)
+ *
+ * - starting_index_kind is one of:
+ * M1: must be -1
+ * chip_id: hardware chip id or -1 for current hw chip
+ * phys_processor_idx:
+ *
+ * __count(offset, bytes, name):
+ * a counter that should be exposed via perf
+ * __field(offset, bytes, name)
+ * a normal field
+ * __array(offset, bytes, name)
+ * an array of bytes
+ *
+ *
+ * @bytes for __count, and __field _must_ be a numeral token
+ * in decimal, not an expression and not in hex.
+ *
+ *
+ * TODO:
+ * - expose secondary index (if any counter ever uses it, only 0xA0
+ * appears to use it right now, and it doesn't have any counters)
+ * - embed versioning info
+ * - include counter descriptions
+ */
+#define REQUEST_NAME dispatch_timebase_by_processor
+#define REQUEST_NUM 0x10
+#define REQUEST_IDX_KIND phys_processor_idx
+#include I(REQUEST_BEGIN)
+REQUEST(__count(0, 8, processor_time_in_timebase_cycles)
+ __field(0x8, 4, hw_processor_id)
+ __field(0xC, 2, owning_part_id)
+ __field(0xE, 1, processor_state)
+ __field(0xF, 1, version)
+ __field(0x10, 4, hw_chip_id)
+ __field(0x14, 4, phys_module_id)
+ __field(0x18, 4, primary_affinity_domain_idx)
+ __field(0x1C, 4, secondary_affinity_domain_idx)
+ __field(0x20, 4, processor_version)
+ __field(0x24, 2, logical_processor_idx)
+ __field(0x26, 2, reserved)
+ __field(0x28, 4, processor_id_register)
+ __field(0x2C, 4, phys_processor_idx)
+)
+#include I(REQUEST_END)
+
+#define REQUEST_NAME system_performance_capabilities
+#define REQUEST_NUM 0x40
+#define REQUEST_IDX_KIND M1
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 1, perf_collect_privileged)
+ __field(0x1, 1 , capability_mask)
+ __array(0x2, 0xE, reserved)
+)
+#include I(REQUEST_END)
+
+#undef M1
+#include "req-gen/_end.h"
diff --git a/arch/powerpc/perf/hv-gpci.c b/arch/powerpc/perf/hv-gpci.c
index 15fc76c..0c8e1c3 100644
--- a/arch/powerpc/perf/hv-gpci.c
+++ b/arch/powerpc/perf/hv-gpci.c
@@ -57,6 +57,11 @@ static struct attribute_group format_group = {
.attrs = format_attrs,
};
+static struct attribute_group event_group = {
+ .name = "events",
+ .attrs = hv_gpci_event_attrs,
+};
+
#define HV_CAPS_ATTR(_name, _format) \
static ssize_t _name##_show(struct device *dev, \
struct device_attribute *attr, \
@@ -102,6 +107,7 @@ static struct attribute_group interface_group = {
static const struct attribute_group *attr_groups[] = {
&format_group,
+ &event_group,
&interface_group,
NULL,
};
@@ -271,6 +277,8 @@ static int hv_gpci_init(void)
unsigned long hret;
struct hv_perf_caps caps;
+ hv_gpci_assert_offsets_correct();
+
if (!firmware_has_feature(FW_FEATURE_LPAR)) {
pr_debug("not a virtualized system, not enabling\n");
return -ENODEV;
diff --git a/arch/powerpc/perf/hv-gpci.h b/arch/powerpc/perf/hv-gpci.h
index b25f460..86ede82 100644
--- a/arch/powerpc/perf/hv-gpci.h
+++ b/arch/powerpc/perf/hv-gpci.h
@@ -42,32 +42,19 @@ struct hv_get_perf_counter_info_params {
*/
#define COUNTER_INFO_VERSION_CURRENT 0x8
-/*
- * These determine the counter_value[] layout and the meaning of starting_index
- * and secondary_index.
- *
- * Unless otherwise noted, @secondary_index is unused and ignored.
- */
-enum counter_info_requests {
-
- /* GENERAL */
-
- /* @starting_index: must be -1 (to refer to the current partition)
- */
- CIR_SYSTEM_PERFORMANCE_CAPABILITIES = 0X40,
+/* capability mask masks. */
+enum {
+ HV_GPCI_CM_GA = (1 << 7),
+ HV_GPCI_CM_EXPANDED = (1 << 6),
+ HV_GPCI_CM_LAB = (1 << 5)
};
-struct cv_system_performance_capabilities {
- /* If != 0, allowed to collect data from other partitions */
- __u8 perf_collect_privileged;
-
- /* These following are only valid if counter_info_version >= 0x3 */
-#define CV_CM_GA (1 << 7)
-#define CV_CM_EXPANDED (1 << 6)
-#define CV_CM_LAB (1 << 5)
- /* remaining bits are reserved */
- __u8 capability_mask;
- __u8 reserved[0xE];
-} __packed;
+#define REQUEST_FILE "../hv-gpci-requests.h"
+#define NAME_LOWER hv_gpci
+#define NAME_UPPER HV_GPCI
+#include "req-gen/perf.h"
+#undef REQUEST_FILE
+#undef NAME_LOWER
+#undef NAME_UPPER
#endif
diff --git a/arch/powerpc/perf/req-gen/_begin.h b/arch/powerpc/perf/req-gen/_begin.h
new file mode 100644
index 0000000..acfb17a
--- /dev/null
+++ b/arch/powerpc/perf/req-gen/_begin.h
@@ -0,0 +1,13 @@
+/* Include paths to be used in interface defining headers */
+#ifndef POWERPC_PERF_REQ_GEN_H_
+#define POWERPC_PERF_REQ_GEN_H_
+
+#define CAT2_STR_(t, s) __stringify(t/s)
+#define CAT2_STR(t, s) CAT2_STR_(t, s)
+#define I(...) __VA_ARGS__
+
+#endif
+
+#define REQ_GEN_PREFIX req-gen
+#define REQUEST_BEGIN CAT2_STR(REQ_GEN_PREFIX, _request-begin.h)
+#define REQUEST_END CAT2_STR(REQ_GEN_PREFIX, _request-end.h)
diff --git a/arch/powerpc/perf/req-gen/_clear.h b/arch/powerpc/perf/req-gen/_clear.h
new file mode 100644
index 0000000..422974f
--- /dev/null
+++ b/arch/powerpc/perf/req-gen/_clear.h
@@ -0,0 +1,5 @@
+
+#undef __field_
+#undef __count_
+#undef __array_
+#undef REQUEST_
diff --git a/arch/powerpc/perf/req-gen/_end.h b/arch/powerpc/perf/req-gen/_end.h
new file mode 100644
index 0000000..8a40698
--- /dev/null
+++ b/arch/powerpc/perf/req-gen/_end.h
@@ -0,0 +1,4 @@
+
+#undef REQ_GEN_PREFIX
+#undef REQUEST_BEGIN
+#undef REQUEST_END
diff --git a/arch/powerpc/perf/req-gen/_request-begin.h b/arch/powerpc/perf/req-gen/_request-begin.h
new file mode 100644
index 0000000..f6d9864
--- /dev/null
+++ b/arch/powerpc/perf/req-gen/_request-begin.h
@@ -0,0 +1,15 @@
+
+#define REQUEST(r_contents) \
+ REQUEST_(REQUEST_NAME, REQUEST_NUM, REQUEST_IDX_KIND, I(r_contents))
+
+#define __field(f_offset, f_bytes, f_name) \
+ __field_(REQUEST_NAME, REQUEST_NUM, REQUEST_IDX_KIND, \
+ f_offset, f_bytes, f_name)
+
+#define __array(f_offset, f_bytes, f_name) \
+ __array_(REQUEST_NAME, REQUEST_NUM, REQUEST_IDX_KIND, \
+ f_offset, f_bytes, f_name)
+
+#define __count(f_offset, f_bytes, f_name) \
+ __count_(REQUEST_NAME, REQUEST_NUM, REQUEST_IDX_KIND, \
+ f_offset, f_bytes, f_name)
diff --git a/arch/powerpc/perf/req-gen/_request-end.h b/arch/powerpc/perf/req-gen/_request-end.h
new file mode 100644
index 0000000..5573be6
--- /dev/null
+++ b/arch/powerpc/perf/req-gen/_request-end.h
@@ -0,0 +1,8 @@
+#undef REQUEST
+#undef __field
+#undef __array
+#undef __count
+
+#undef REQUEST_NAME
+#undef REQUEST_NUM
+#undef REQUEST_IDX_KIND
diff --git a/arch/powerpc/perf/req-gen/perf.h b/arch/powerpc/perf/req-gen/perf.h
new file mode 100644
index 0000000..591d663
--- /dev/null
+++ b/arch/powerpc/perf/req-gen/perf.h
@@ -0,0 +1,155 @@
+#ifndef LINUX_POWERPC_PERF_REQ_GEN_PERF_H_
+#define LINUX_POWERPC_PERF_REQ_GEN_PERF_H_
+
+#include <linux/perf_event.h>
+
+#ifndef REQUEST_FILE
+#error "REQUEST_FILE must be defined before including"
+#endif
+
+#ifndef NAME_LOWER
+#error "NAME_LOWER must be defined before including"
+#endif
+
+#ifndef NAME_UPPER
+#error "NAME_UPPER must be defined before including"
+#endif
+
+#define BE_TYPE_b1 __u8
+#define BE_TYPE_b2 __be16
+#define BE_TYPE_b4 __be32
+#define BE_TYPE_b8 __be64
+
+#define BYTES_TO_BE_TYPE(bytes) \
+ BE_TYPE_b##bytes
+
+#define CAT2_(a, b) a ## b
+#define CAT2(a, b) CAT2_(a, b)
+#define CAT3_(a, b, c) a ## b ## c
+#define CAT3(a, b, c) CAT3_(a, b, c)
+
+/*
+ * enumerate the request values as
+ * <NAME_UPPER>_<request name> = <request value>
+ */
+#define REQUEST_VALUE__(name_upper, r_name) name_upper ## _ ## r_name
+#define REQUEST_VALUE_(name_upper, r_name) REQUEST_VALUE__(name_upper, r_name)
+#define REQUEST_VALUE(r_name) REQUEST_VALUE_(NAME_UPPER, r_name)
+
+#include "_clear.h"
+#define REQUEST_(r_name, r_value, r_idx_1, r_fields) \
+ REQUEST_VALUE(r_name) = r_value,
+enum CAT2(NAME_LOWER, _requests) {
+#include REQUEST_FILE
+};
+
+/*
+ * For each request:
+ * struct <NAME_LOWER>_<request name> {
+ * r_fields
+ * };
+ */
+#include "_clear.h"
+#define STRUCT_NAME__(name_lower, r_name) name_lower ## _ ## r_name
+#define STRUCT_NAME_(name_lower, r_name) STRUCT_NAME__(name_lower, r_name)
+#define STRUCT_NAME(r_name) STRUCT_NAME_(NAME_LOWER, r_name)
+#define REQUEST_(r_name, r_value, r_idx_1, r_fields) \
+struct STRUCT_NAME(r_name) { \
+ r_fields \
+};
+#define __field_(r_name, r_value, r_idx_1, f_offset, f_bytes, f_name) \
+ BYTES_TO_BE_TYPE(f_bytes) f_name;
+#define __count_(r_name, r_value, r_idx_1, f_offset, f_bytes, f_name) \
+ __field_(r_name, r_value, r_idx_1, f_offset, f_bytes, f_name)
+#define __array_(r_name, r_value, r_idx_1, a_offset, a_bytes, a_name) \
+ __u8 a_name[a_bytes];
+
+#include REQUEST_FILE
+
+/*
+ * Generate a check of the field offsets
+ * <NAME_LOWER>_assert_offsets_correct()
+ */
+#include "_clear.h"
+#define REQUEST_(r_name, r_value, index, r_fields) \
+r_fields
+#define __field_(r_name, r_value, r_idx_1, f_offset, f_size, f_name) \
+ BUILD_BUG_ON(offsetof(struct STRUCT_NAME(r_name), f_name) != f_offset);
+#define __count_(r_name, r_value, r_idx_1, c_offset, c_size, c_name) \
+ __field_(r_name, r_value, r_idx_1, c_offset, c_size, c_name)
+#define __array_(r_name, r_value, r_idx_1, a_offset, a_size, a_name) \
+ __field_(r_name, r_value, r_idx_1, a_offset, a_size, a_name)
+
+static inline void CAT2(NAME_LOWER, _assert_offsets_correct)(void)
+{
+#include REQUEST_FILE
+}
+
+/*
+ * Generate event attributes:
+ * PMU_EVENT_ATTR_STRING(<request name>_<field name>,
+ * <NAME_LOWER>_event_attr_<request name>_<field name>,
+ * "request=<request value>"
+ * "starting_index=<starting index type>"
+ * "counter_info_version=CURRENT_COUNTER_INFO_VERSION"
+ * "length=<f_size>"
+ * "offset=<f_offset>")
+ *
+ * TODO: counter_info_version may need to vary, we should interperate the
+ * value to some extent
+ */
+#define EVENT_ATTR_NAME__(name, r_name, c_name) \
+ name ## _event_attr_ ## r_name ## _ ## c_name
+#define EVENT_ATTR_NAME_(name, r_name, c_name) \
+ EVENT_ATTR_NAME__(name, r_name, c_name)
+#define EVENT_ATTR_NAME(r_name, c_name) \
+ EVENT_ATTR_NAME_(NAME_LOWER, r_name, c_name)
+
+#include "_clear.h"
+#define __field_(r_name, r_value, r_idx_1, f_offset, f_size, f_name)
+#define __array_(r_name, r_value, r_idx_1, a_offset, a_size, a_name)
+#define __count_(r_name, r_value, r_idx_1, c_offset, c_size, c_name) \
+PMU_EVENT_ATTR_STRING( \
+ CAT3(r_name, _, c_name), \
+ EVENT_ATTR_NAME(r_name, c_name), \
+ "request=" __stringify(r_value) "," \
+ "starting_index=" __stringify(r_idx_1) "," \
+ "counter_info_version=" \
+ __stringify(COUNTER_INFO_VERSION_CURRENT) "," \
+ "length=" #c_size "," \
+ "offset=" #c_offset)
+#define REQUEST_(r_name, r_value, r_idx_1, r_fields) \
+ r_fields
+
+#include REQUEST_FILE
+
+/*
+ * Define event attribute array
+ * static struct attribute *hv_gpci_event_attrs[] = {
+ * &<NAME_LOWER>_event_attr_<request name>_<field name>.attr,
+ * };
+ */
+#include "_clear.h"
+#define __field_(r_name, r_value, r_idx_1, f_offset, f_size, f_name)
+#define __count_(r_name, r_value, r_idx_1, c_offset, c_size, c_name) \
+ &EVENT_ATTR_NAME(r_name, c_name).attr.attr,
+#define __array_(r_name, r_value, r_idx_1, a_offset, a_size, a_name)
+#define REQUEST_(r_name, r_value, r_idx_1, r_fields) \
+ r_fields
+
+static __maybe_unused struct attribute *hv_gpci_event_attrs[] = {
+#include REQUEST_FILE
+ NULL
+};
+
+/* cleanup */
+#include "_clear.h"
+#undef EVENT_ATTR_NAME
+#undef EVENT_ATTR_NAME_
+#undef BIT_NAME
+#undef BIT_NAME_
+#undef STRUCT_NAME
+#undef REQUEST_VALUE
+#undef REQUEST_VALUE_
+
+#endif
--
1.7.9.5
^ permalink raw reply related
* [PATCH v3 07/10] powerpc/perf/hv-gpci: add the remaining gpci requests
From: Sukadev Bhattiprolu @ 2014-09-17 18:52 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, ak, Jiri Olsa, peterz, eranian,
Paul Mackerras
Cc: Michael Ellerman, linux-kernel, dev, linuxppc-dev,
Anshuman Khandual
In-Reply-To: <1410979923-7722-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
Add the remaining gpci requests that contain counters suitable for use
by perf. Omit those that don't contain any counters (but note their
ommision).
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
arch/powerpc/perf/hv-gpci-requests.h | 183 ++++++++++++++++++++++++++++++++++
1 file changed, 183 insertions(+)
diff --git a/arch/powerpc/perf/hv-gpci-requests.h b/arch/powerpc/perf/hv-gpci-requests.h
index 0dfc4d9..a908b08 100644
--- a/arch/powerpc/perf/hv-gpci-requests.h
+++ b/arch/powerpc/perf/hv-gpci-requests.h
@@ -65,6 +65,33 @@ REQUEST(__count(0, 8, processor_time_in_timebase_cycles)
)
#include I(REQUEST_END)
+#define REQUEST_NAME entitled_capped_uncapped_donated_idle_timebase_by_partition
+#define REQUEST_NUM 0x20
+#define REQUEST_IDX_KIND sibling_part_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 8, partition_id)
+ __count(0x8, 8, entitled_cycles)
+ __count(0x10, 8, consumed_capped_cycles)
+ __count(0x18, 8, consumed_uncapped_cycles)
+ __count(0x20, 8, cycles_donated)
+ __count(0x28, 8, purr_idle_cycles)
+)
+#include I(REQUEST_END)
+
+/*
+ * Not avaliable for counter_info_version >= 0x8, use
+ * run_instruction_cycles_by_partition(0x100) instead.
+ */
+#define REQUEST_NAME run_instructions_run_cycles_by_partition
+#define REQUEST_NUM 0x30
+#define REQUEST_IDX_KIND sibling_part_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 8, partition_id)
+ __count(0x8, 8, instructions_completed)
+ __count(0x10, 8, cycles)
+)
+#include I(REQUEST_END)
+
#define REQUEST_NAME system_performance_capabilities
#define REQUEST_NUM 0x40
#define REQUEST_IDX_KIND M1
@@ -75,5 +102,161 @@ REQUEST(__field(0, 1, perf_collect_privileged)
)
#include I(REQUEST_END)
+#define REQUEST_NAME processor_bus_utilization_abc_links
+#define REQUEST_NUM 0x50
+#define REQUEST_IDX_KIND hw_chip_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 4, hw_chip_id)
+ __array(0x4, 0xC, reserved1)
+ __count(0x10, 8, total_link_cycles)
+ __count(0x18, 8, idle_cycles_for_a_link)
+ __count(0x20, 8, idle_cycles_for_b_link)
+ __count(0x28, 8, idle_cycles_for_c_link)
+ __array(0x30, 0x20, reserved2)
+)
+#include I(REQUEST_END)
+
+#define REQUEST_NAME processor_bus_utilization_wxyz_links
+#define REQUEST_NUM 0x60
+#define REQUEST_IDX_KIND hw_chip_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 4, hw_chip_id)
+ __array(0x4, 0xC, reserved1)
+ __count(0x10, 8, total_link_cycles)
+ __count(0x18, 8, idle_cycles_for_w_link)
+ __count(0x20, 8, idle_cycles_for_x_link)
+ __count(0x28, 8, idle_cycles_for_y_link)
+ __count(0x30, 8, idle_cycles_for_z_link)
+ __array(0x38, 0x28, reserved2)
+)
+#include I(REQUEST_END)
+
+#define REQUEST_NAME processor_bus_utilization_gx_links
+#define REQUEST_NUM 0x70
+#define REQUEST_IDX_KIND hw_chip_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 4, hw_chip_id)
+ __array(0x4, 0xC, reserved1)
+ __count(0x10, 8, gx0_in_address_cycles)
+ __count(0x18, 8, gx0_in_data_cycles)
+ __count(0x20, 8, gx0_in_retries)
+ __count(0x28, 8, gx0_in_bus_cycles)
+ __count(0x30, 8, gx0_in_cycles_total)
+ __count(0x38, 8, gx0_out_address_cycles)
+ __count(0x40, 8, gx0_out_data_cycles)
+ __count(0x48, 8, gx0_out_retries)
+ __count(0x50, 8, gx0_out_bus_cycles)
+ __count(0x58, 8, gx0_out_cycles_total)
+ __count(0x60, 8, gx1_in_address_cycles)
+ __count(0x68, 8, gx1_in_data_cycles)
+ __count(0x70, 8, gx1_in_retries)
+ __count(0x78, 8, gx1_in_bus_cycles)
+ __count(0x80, 8, gx1_in_cycles_total)
+ __count(0x88, 8, gx1_out_address_cycles)
+ __count(0x90, 8, gx1_out_data_cycles)
+ __count(0x98, 8, gx1_out_retries)
+ __count(0xA0, 8, gx1_out_bus_cycles)
+ __count(0xA8, 8, gx1_out_cycles_total)
+)
+#include I(REQUEST_END)
+
+#define REQUEST_NAME processor_bus_utilization_mc_links
+#define REQUEST_NUM 0x80
+#define REQUEST_IDX_KIND hw_chip_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 4, hw_chip_id)
+ __array(0x4, 0xC, reserved1)
+ __count(0x10, 8, mc0_frames)
+ __count(0x18, 8, mc0_reads)
+ __count(0x20, 8, mc0_write)
+ __count(0x28, 8, mc0_total_cycles)
+ __count(0x30, 8, mc1_frames)
+ __count(0x38, 8, mc1_reads)
+ __count(0x40, 8, mc1_writes)
+ __count(0x48, 8, mc1_total_cycles)
+)
+#include I(REQUEST_END)
+
+/* Processor_config (0x90) skipped, no counters */
+/* Current_processor_frequency (0x91) skipped, no counters */
+
+#define REQUEST_NAME processor_core_utilization
+#define REQUEST_NUM 0x94
+#define REQUEST_IDX_KIND phys_processor_idx
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 4, phys_processor_idx)
+ __field(0x4, 4, hw_processor_id)
+ __count(0x8, 8, cycles_across_any_thread)
+ __count(0x10, 8, timebase_at_collection)
+ __count(0x18, 8, purr_cycles)
+ __count(0x20, 8, sum_of_cycles_across_all_threads)
+ __count(0x28, 8, instructions_completed)
+)
+#include I(REQUEST_END)
+
+/* Processor_core_power_mode (0x95) skipped, no counters */
+/* Affinity_domain_information_by_virtual_processor (0xA0) skipped,
+ * no counters */
+/* Affinity_domain_information_by_domain (0xB0) skipped, no counters */
+/* Affinity_domain_information_by_partition (0xB1) skipped, no counters */
+/* Physical_memory_info (0xC0) skipped, no counters */
+/* Processor_bus_topology (0xD0) skipped, no counters */
+
+#define REQUEST_NAME partition_hypervisor_queuing_times
+#define REQUEST_NUM 0xE0
+#define REQUEST_IDX_KIND partition_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 2, partition_id)
+ __array(0x2, 6, reserved1)
+ __count(0x8, 8, time_waiting_for_entitlement)
+ __count(0x10, 8, times_waited_for_entitlement)
+ __count(0x18, 8, time_waiting_for_phys_processor)
+ __count(0x20, 8, times_waited_for_phys_processor)
+ __count(0x28, 8, dispatches_on_home_core)
+ __count(0x30, 8, dispatches_on_home_primary_affinity_domain)
+ __count(0x38, 8, dispatches_on_home_secondary_affinity_domain)
+ __count(0x40, 8, dispatches_off_home_secondary_affinity_domain)
+ __count(0x48, 8, dispatches_on_dedicated_processor_donating_cycles)
+)
+#include I(REQUEST_END)
+
+#define REQUEST_NAME system_hypervisor_times
+#define REQUEST_NUM 0xF0
+#define REQUEST_IDX_KIND M1
+#include I(REQUEST_BEGIN)
+REQUEST(__count(0, 8, time_spent_to_dispatch_virtual_processors)
+ __count(0x8, 8, time_spent_processing_virtual_processor_timers)
+ __count(0x10, 8, time_spent_managing_partitions_over_entitlement)
+ __count(0x18, 8, time_spent_on_system_managment)
+)
+#include I(REQUEST_END)
+
+#define REQUEST_NAME system_tlbie_count_and_time
+#define REQUEST_NUM 0xF4
+#define REQUEST_IDX_KIND M1
+#include I(REQUEST_BEGIN)
+REQUEST(__count(0, 8, tlbie_instructions_issued)
+ /*
+ * FIXME: The spec says the offset here is 0x10, which I suspect
+ * is wrong.
+ */
+ __count(0x8, 8, time_spent_issuing_tlbies)
+)
+#include I(REQUEST_END)
+
+#define REQUEST_NAME partition_instruction_count_and_time
+#define REQUEST_NUM 0x100
+#define REQUEST_IDX_KIND partition_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 2, partition_id)
+ __array(0x2, 0x6, reserved1)
+ __count(0x8, 8, instructions_performed)
+ __count(0x10, 8, time_collected)
+)
+#include I(REQUEST_END)
+
+/* set_mmcrh (0x80001000) skipped, no counters */
+/* retrieve_hpmcx (0x80002000) skipped, no counters */
+
#undef M1
#include "req-gen/_end.h"
--
1.7.9.5
^ permalink raw reply related
* [PATCH v3 05/10] perf: add PMU_EVENT_ATTR_STRING() helper
From: Sukadev Bhattiprolu @ 2014-09-17 18:51 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, ak, Jiri Olsa, peterz, eranian,
Paul Mackerras
Cc: Michael Ellerman, linux-kernel, dev, linuxppc-dev,
Anshuman Khandual
In-Reply-To: <1410979923-7722-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
Helper for constructing static struct perf_pmu_events_attr s.
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
include/linux/perf_event.h | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index cef4a56..32190c5 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -891,6 +891,13 @@ static struct perf_pmu_events_attr _var = { \
.id = _id, \
};
+#define PMU_EVENT_ATTR_STRING(_name, _var, _value) \
+static struct perf_pmu_events_attr _var = { \
+ .attr = __ATTR(_name, 0444, perf_event_sysfs_show, NULL), \
+ .event_str = _value, \
+};
+
+
#define PMU_FORMAT_ATTR(_name, _format) \
static ssize_t \
_name##_show(struct device *dev, \
--
1.7.9.5
^ permalink raw reply related
* [PATCH v3 04/10] powerpc/perf/hv-24x7: parse catalog and populate sysfs with events
From: Sukadev Bhattiprolu @ 2014-09-17 18:51 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, ak, Jiri Olsa, peterz, eranian,
Paul Mackerras
Cc: Michael Ellerman, linux-kernel, dev, linuxppc-dev,
Anshuman Khandual
In-Reply-To: <1410979923-7722-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
Retrieves and parses the 24x7 catalog on POWER systems that supply it
(right now, only POWER 8). Events are exposed via sysfs in the standard
fashion, and are all parameterized.
Catalog is (at the moment) only parsed on boot. It needs re-parsing
when a some hypervisor events occur. At that point we'll also need to
prevent old events from continuing to function (counter that is passed
in via spare space in the config values?).
Changelog[v6]
[Sukadev Bhattiprolu] Use PHYS and VCPU in place of PHYSICAL and
VIRTUAL_PROCESSOR to shorten the names of the domains and hence,
events;
Changelog[v2]
[Joe Perches, David Laight] Use beNN_to_cpu() instead of guessing
the size from type.
Use kmem_cache_free() to free page allocated with kmem_cache_alloc().
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
arch/powerpc/perf/hv-24x7-catalog.h | 25 ++
arch/powerpc/perf/hv-24x7-domains.h | 28 ++
arch/powerpc/perf/hv-24x7.c | 787 ++++++++++++++++++++++++++++++++++-
arch/powerpc/perf/hv-24x7.h | 12 +-
4 files changed, 838 insertions(+), 14 deletions(-)
create mode 100644 arch/powerpc/perf/hv-24x7-domains.h
diff --git a/arch/powerpc/perf/hv-24x7-catalog.h b/arch/powerpc/perf/hv-24x7-catalog.h
index 21b19dd..69e2e1f 100644
--- a/arch/powerpc/perf/hv-24x7-catalog.h
+++ b/arch/powerpc/perf/hv-24x7-catalog.h
@@ -30,4 +30,29 @@ struct hv_24x7_catalog_page_0 {
__u8 reserved6[2];
} __packed;
+struct hv_24x7_event_data {
+ __be16 length; /* in bytes, must be a multiple of 16 */
+ __u8 reserved1[2];
+ __u8 domain; /* Chip = 1, Core = 2 */
+ __u8 reserved2[1];
+ __be16 event_group_record_offs; /* in bytes, must be 8 byte aligned */
+ __be16 event_group_record_len; /* in bytes */
+
+ /* in bytes, offset from event_group_record */
+ __be16 event_counter_offs;
+
+ /* verified_state, unverified_state, caveat_state, broken_state, ... */
+ __be32 flags;
+
+ __be16 primary_group_ix;
+ __be16 group_count;
+ __be16 event_name_len;
+ __u8 remainder[];
+ /* __u8 event_name[event_name_len - 2]; */
+ /* __be16 event_description_len; */
+ /* __u8 event_desc[event_description_len - 2]; */
+ /* __be16 detailed_desc_len; */
+ /* __u8 detailed_desc[detailed_desc_len - 2]; */
+} __packed;
+
#endif
diff --git a/arch/powerpc/perf/hv-24x7-domains.h b/arch/powerpc/perf/hv-24x7-domains.h
new file mode 100644
index 0000000..49c1efd
--- /dev/null
+++ b/arch/powerpc/perf/hv-24x7-domains.h
@@ -0,0 +1,28 @@
+
+/*
+ * DOMAIN(name, num, index_kind, is_physical)
+ *
+ * @name: An all caps token, suitable for use in generating an enum
+ * member and appending to an event name in sysfs.
+ *
+ * @num: The number corresponding to the domain as given in
+ * documentation. We assume the catalog domain and the hcall
+ * domain have the same numbering (so far they do), but this
+ * may need to be changed in the future.
+ *
+ * @index_kind: A stringifiable token describing the meaning of the index
+ * within the given domain. Must fit the parsing rules of the
+ * perf sysfs api.
+ *
+ * @is_physical: True if the domain is physical, false otherwise (if virtual).
+ *
+ * Note: The terms PHYS_CHIP, PHYS_CORE, VCPU correspond to physical chip,
+ * physical core and virtual processor in 24x7 Counters specifications.
+ */
+
+DOMAIN(PHYS_CHIP, 0x01, chip, true)
+DOMAIN(PHYS_CORE, 0x02, core, true)
+DOMAIN(VCPU_HOME_CORE, 0x03, vcpu, false)
+DOMAIN(VCPU_HOME_CHIP, 0x04, vcpu, false)
+DOMAIN(VCPU_HOME_NODE, 0x05, vcpu, false)
+DOMAIN(VCPU_REMOTE_NODE, 0x06, vcpu, false)
diff --git a/arch/powerpc/perf/hv-24x7.c b/arch/powerpc/perf/hv-24x7.c
index 9427ef7..f13e2ea 100644
--- a/arch/powerpc/perf/hv-24x7.c
+++ b/arch/powerpc/perf/hv-24x7.c
@@ -13,16 +13,80 @@
#define pr_fmt(fmt) "hv-24x7: " fmt
#include <linux/perf_event.h>
+#include <linux/rbtree.h>
#include <linux/module.h>
#include <linux/slab.h>
+#include <linux/vmalloc.h>
+
#include <asm/firmware.h>
#include <asm/hvcall.h>
#include <asm/io.h>
+#include <linux/byteorder/generic.h>
#include "hv-24x7.h"
#include "hv-24x7-catalog.h"
#include "hv-common.h"
+static const char *domain_to_index_string(unsigned domain)
+{
+ switch (domain) {
+#define DOMAIN(n, v, x, c) \
+ case HV_PERF_DOMAIN_##n: \
+ return #x;
+#include "hv-24x7-domains.h"
+#undef DOMAIN
+ default:
+ WARN(1, "unknown domain %d\n", domain);
+ return "UNKNOWN_DOMAIN_INDEX_STRING";
+ }
+}
+
+static const char *event_domain_suffix(unsigned domain)
+{
+ switch (domain) {
+#define DOMAIN(n, v, x, c) \
+ case HV_PERF_DOMAIN_##n: \
+ return "__" #n;
+#include "hv-24x7-domains.h"
+#undef DOMAIN
+ default:
+ WARN(1, "unknown domain %d\n", domain);
+ return "__UNKNOWN_DOMAIN_SUFFIX";
+ }
+}
+
+static bool domain_is_valid(unsigned domain)
+{
+ switch (domain) {
+#define DOMAIN(n, v, x, c) \
+ case HV_PERF_DOMAIN_##n: \
+ /* fall through */
+#include "hv-24x7-domains.h"
+#undef DOMAIN
+ return true;
+ default:
+ return false;
+ }
+}
+
+static bool is_physical_domain(unsigned domain)
+{
+ switch (domain) {
+#define DOMAIN(n, v, x, c) \
+ case HV_PERF_DOMAIN_##n: \
+ return c;
+#include "hv-24x7-domains.h"
+#undef DOMAIN
+ default:
+ return false;
+ }
+}
+
+static bool catalog_entry_domain_is_valid(unsigned domain)
+{
+ return is_physical_domain(domain);
+}
+
/*
* TODO: Merging events:
* - Think of the hcall as an interface to a 4d array of counters:
@@ -73,8 +137,115 @@ static struct attribute_group format_group = {
.attrs = format_attrs,
};
+static struct attribute_group event_group = {
+ .name = "events",
+ /* .attrs is set in init */
+};
+
+static struct attribute_group event_desc_group = {
+ .name = "event_descs",
+ /* .attrs is set in init */
+};
+
+static struct attribute_group event_long_desc_group = {
+ .name = "event_long_descs",
+ /* .attrs is set in init */
+};
+
static struct kmem_cache *hv_page_cache;
+static char *event_name(struct hv_24x7_event_data *ev, int *len)
+{
+ *len = be16_to_cpu(ev->event_name_len) - 2;
+ return (char *)ev->remainder;
+}
+
+static char *event_desc(struct hv_24x7_event_data *ev, int *len)
+{
+ unsigned nl = be16_to_cpu(ev->event_name_len);
+ __be16 *desc_len = (__be16 *)(ev->remainder + nl - 2);
+ *len = be16_to_cpu(*desc_len) - 2;
+ return (char *)ev->remainder + nl;
+}
+
+static char *event_long_desc(struct hv_24x7_event_data *ev, int *len)
+{
+ unsigned nl = be16_to_cpu(ev->event_name_len);
+ __be16 *desc_len_ = (__be16 *)(ev->remainder + nl - 2);
+ unsigned desc_len = be16_to_cpu(*desc_len_);
+ __be16 *long_desc_len = (__be16 *)(ev->remainder + nl + desc_len - 2);
+ *len = be16_to_cpu(*long_desc_len) - 2;
+ return (char *)ev->remainder + nl + desc_len;
+}
+
+static bool event_fixed_portion_is_within(struct hv_24x7_event_data *ev,
+ void *end)
+{
+ void *start = ev;
+
+ return (start + offsetof(struct hv_24x7_event_data, remainder)) < end;
+}
+
+/*
+ * Things we don't check:
+ * - padding for desc, name, and long/detailed desc is required to be '\0'
+ * bytes.
+ *
+ * Return NULL if we pass end,
+ * Otherwise return the address of the byte just following the event.
+ */
+static void *event_end(struct hv_24x7_event_data *ev, void *end)
+{
+ void *start = ev;
+ __be16 *dl_, *ldl_;
+ unsigned dl, ldl;
+ unsigned nl = be16_to_cpu(ev->event_name_len);
+
+ if (nl < 2) {
+ pr_debug("%s: name length too short: %d", __func__, nl);
+ return NULL;
+ }
+
+ if (start + nl > end) {
+ pr_debug("%s: start=%p + nl=%u > end=%p",
+ __func__, start, nl, end);
+ return NULL;
+ }
+
+ dl_ = (__be16 *)(ev->remainder + nl - 2);
+ if (!IS_ALIGNED((uintptr_t)dl_, 2))
+ pr_warn("desc len not aligned %p", dl_);
+ dl = be16_to_cpu(*dl_);
+ if (dl < 2) {
+ pr_debug("%s: desc len too short: %d", __func__, dl);
+ return NULL;
+ }
+
+ if (start + nl + dl > end) {
+ pr_debug("%s: (start=%p + nl=%u + dl=%u)=%p > end=%p",
+ __func__, start, nl, dl, start + nl + dl, end);
+ return NULL;
+ }
+
+ ldl_ = (__be16 *)(ev->remainder + nl + dl - 2);
+ if (!IS_ALIGNED((uintptr_t)ldl_, 2))
+ pr_warn("long desc len not aligned %p", ldl_);
+ ldl = be16_to_cpu(*ldl_);
+ if (ldl < 2) {
+ pr_debug("%s: long desc len too short (ldl=%u)",
+ __func__, ldl);
+ return NULL;
+ }
+
+ if (start + nl + dl + ldl > end) {
+ pr_debug("%s: start=%p + nl=%u + dl=%u + ldl=%u > end=%p",
+ __func__, start, nl, dl, ldl, end);
+ return NULL;
+ }
+
+ return start + nl + dl + ldl;
+}
+
static unsigned long h_get_24x7_catalog_page_(unsigned long phys_4096,
unsigned long version,
unsigned long index)
@@ -97,6 +268,605 @@ static unsigned long h_get_24x7_catalog_page(char page[],
version, index);
}
+static unsigned core_domains[] = {
+ HV_PERF_DOMAIN_PHYS_CORE,
+ HV_PERF_DOMAIN_VCPU_HOME_CORE,
+ HV_PERF_DOMAIN_VCPU_HOME_CHIP,
+ HV_PERF_DOMAIN_VCPU_HOME_NODE,
+ HV_PERF_DOMAIN_VCPU_REMOTE_NODE,
+};
+/* chip event data always yeilds a single event, core yeilds multiple */
+#define MAX_EVENTS_PER_EVENT_DATA ARRAY_SIZE(core_domains)
+
+static char *event_fmt(struct hv_24x7_event_data *event, unsigned domain)
+{
+ const char *lpar;
+
+ if (is_physical_domain(domain))
+ lpar = "0x0";
+ else
+ lpar = "sibling_guest_id";
+
+ return kasprintf(GFP_KERNEL,
+ "domain=0x%x,offset=0x%x,starting_index=%s,lpar=%s",
+ domain,
+ be16_to_cpu(event->event_counter_offs) +
+ be16_to_cpu(event->event_group_record_offs),
+ domain_to_index_string(domain),
+ lpar);
+}
+
+/* Avoid trusting fw to NUL terminate strings */
+static char *memdup_to_str(char *maybe_str, int max_len, gfp_t gfp)
+{
+ return kasprintf(gfp, "%.*s", max_len, maybe_str);
+}
+
+static ssize_t device_show_string(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ struct dev_ext_attribute *d;
+
+ d = container_of(attr, struct dev_ext_attribute, attr);
+ return sprintf(buf, "%s\n", (char *)d->var);
+}
+
+static struct attribute *device_str_attr_create_(char *name, char *str)
+{
+ struct dev_ext_attribute *attr = kzalloc(sizeof(*attr), GFP_KERNEL);
+
+ if (!attr)
+ return NULL;
+
+ attr->var = str;
+ attr->attr.attr.name = name;
+ attr->attr.attr.mode = 0444;
+ attr->attr.show = device_show_string;
+ return &attr->attr.attr;
+}
+
+static struct attribute *device_str_attr_create(char *name, int name_max,
+ int name_nonce,
+ char *str, size_t str_max)
+{
+ char *n;
+ char *s = memdup_to_str(str, str_max, GFP_KERNEL);
+ struct attribute *a;
+
+ if (!s)
+ return NULL;
+
+ if (!name_nonce)
+ n = kasprintf(GFP_KERNEL, "%.*s", name_max, name);
+ else
+ n = kasprintf(GFP_KERNEL, "%.*s__%d", name_max, name,
+ name_nonce);
+ if (!n)
+ goto out_s;
+
+ a = device_str_attr_create_(n, s);
+ if (!a)
+ goto out_n;
+
+ return a;
+out_n:
+ kfree(n);
+out_s:
+ kfree(s);
+ return NULL;
+}
+
+static void device_str_attr_destroy(struct attribute *attr)
+{
+ struct dev_ext_attribute *d;
+
+ d = container_of(attr, struct dev_ext_attribute, attr.attr);
+ kfree(d->var);
+ kfree(d->attr.attr.name);
+ kfree(d);
+}
+
+static struct attribute *event_to_attr(unsigned ix,
+ struct hv_24x7_event_data *event,
+ unsigned domain,
+ int nonce)
+{
+ int event_name_len;
+ char *ev_name, *a_ev_name, *val;
+ const char *ev_suffix;
+ struct attribute *attr;
+
+ if (!domain_is_valid(domain)) {
+ pr_warn("catalog event %u has invalid domain %u\n",
+ ix, domain);
+ return NULL;
+ }
+
+ val = event_fmt(event, domain);
+ if (!val)
+ return NULL;
+
+ ev_suffix = event_domain_suffix(domain);
+ ev_name = event_name(event, &event_name_len);
+ if (!nonce)
+ a_ev_name = kasprintf(GFP_KERNEL, "%.*s%s",
+ (int)event_name_len, ev_name, ev_suffix);
+ else
+ a_ev_name = kasprintf(GFP_KERNEL, "%.*s%s__%d",
+ (int)event_name_len, ev_name, ev_suffix, nonce);
+
+
+ if (!a_ev_name)
+ goto out_val;
+
+ attr = device_str_attr_create_(a_ev_name, val);
+ if (!attr)
+ goto out_name;
+
+ return attr;
+out_name:
+ kfree(a_ev_name);
+out_val:
+ kfree(val);
+ return NULL;
+}
+
+static struct attribute *event_to_desc_attr(struct hv_24x7_event_data *event,
+ int nonce)
+{
+ int nl, dl;
+ char *name = event_name(event, &nl);
+ char *desc = event_desc(event, &dl);
+
+ /* If there isn't a description, don't create the sysfs file */
+ if (!dl)
+ return NULL;
+
+ return device_str_attr_create(name, nl, nonce, desc, dl);
+}
+
+static struct attribute *
+event_to_long_desc_attr(struct hv_24x7_event_data *event, int nonce)
+{
+ int nl, dl;
+ char *name = event_name(event, &nl);
+ char *desc = event_long_desc(event, &dl);
+
+ /* If there isn't a description, don't create the sysfs file */
+ if (!dl)
+ return NULL;
+
+ return device_str_attr_create(name, nl, nonce, desc, dl);
+}
+
+static ssize_t event_data_to_attrs(unsigned ix, struct attribute **attrs,
+ struct hv_24x7_event_data *event, int nonce)
+{
+ unsigned i;
+
+ switch (event->domain) {
+ case HV_PERF_DOMAIN_PHYS_CHIP:
+ *attrs = event_to_attr(ix, event, event->domain, nonce);
+ return 1;
+ case HV_PERF_DOMAIN_PHYS_CORE:
+ for (i = 0; i < ARRAY_SIZE(core_domains); i++) {
+ attrs[i] = event_to_attr(ix, event, core_domains[i],
+ nonce);
+ if (!attrs[i]) {
+ pr_warn("catalog event %u: individual attr %u "
+ "creation failure\n", ix, i);
+ for (; i; i--)
+ device_str_attr_destroy(attrs[i - 1]);
+ return -1;
+ }
+ }
+ return i;
+ default:
+ pr_warn("catalog event %u: domain %u is not allowed in the "
+ "catalog\n", ix, event->domain);
+ return -1;
+ }
+}
+
+static size_t event_to_attr_ct(struct hv_24x7_event_data *event)
+{
+ switch (event->domain) {
+ case HV_PERF_DOMAIN_PHYS_CHIP:
+ return 1;
+ case HV_PERF_DOMAIN_PHYS_CORE:
+ return ARRAY_SIZE(core_domains);
+ default:
+ return 0;
+ }
+}
+
+static unsigned long vmalloc_to_phys(void *v)
+{
+ struct page *p = vmalloc_to_page(v);
+
+ BUG_ON(!p);
+ return page_to_phys(p) + offset_in_page(v);
+}
+
+/* */
+struct event_uniq {
+ struct rb_node node;
+ const char *name;
+ int nl;
+ unsigned ct;
+ unsigned domain;
+};
+
+static int memord(const void *d1, size_t s1, const void *d2, size_t s2)
+{
+ if (s1 < s2)
+ return 1;
+ if (s2 > s1)
+ return -1;
+
+ return memcmp(d1, d2, s1);
+}
+
+static int ev_uniq_ord(const void *v1, size_t s1, unsigned d1, const void *v2,
+ size_t s2, unsigned d2)
+{
+ int r = memord(v1, s1, v2, s2);
+
+ if (r)
+ return r;
+ if (d1 > d2)
+ return 1;
+ if (d2 > d1)
+ return -1;
+ return 0;
+}
+
+static int event_uniq_add(struct rb_root *root, const char *name, int nl,
+ unsigned domain)
+{
+ struct rb_node **new = &(root->rb_node), *parent = NULL;
+ struct event_uniq *data;
+
+ /* Figure out where to put new node */
+ while (*new) {
+ struct event_uniq *it;
+ int result;
+
+ it = container_of(*new, struct event_uniq, node);
+ result = ev_uniq_ord(name, nl, domain, it->name, it->nl,
+ it->domain);
+
+ parent = *new;
+ if (result < 0)
+ new = &((*new)->rb_left);
+ else if (result > 0)
+ new = &((*new)->rb_right);
+ else {
+ it->ct++;
+ pr_info("found a duplicate event %.*s, ct=%u\n", nl,
+ name, it->ct);
+ return it->ct;
+ }
+ }
+
+ data = kmalloc(sizeof(*data), GFP_KERNEL);
+ if (!data)
+ return -ENOMEM;
+
+ *data = (struct event_uniq) {
+ .name = name,
+ .nl = nl,
+ .ct = 0,
+ .domain = domain,
+ };
+
+ /* Add new node and rebalance tree. */
+ rb_link_node(&data->node, parent, new);
+ rb_insert_color(&data->node, root);
+
+ /* data->ct */
+ return 0;
+}
+
+static void event_uniq_destroy(struct rb_root *root)
+{
+ /*
+ * the strings we point to are in the giant block of memory filled by
+ * the catalog, and are freed seperately.
+ */
+ struct event_uniq *pos, *n;
+
+ rbtree_postorder_for_each_entry_safe(pos, n, root, node)
+ kfree(pos);
+}
+
+
+/*
+ * ensure the event structure's sizes are self consistent and don't cause us to
+ * read outside of the event
+ *
+ * On success, return the event length in bytes.
+ * Otherwise, return -1 (and print as appropriate).
+ */
+static ssize_t catalog_event_len_validate(struct hv_24x7_event_data *event,
+ size_t event_idx,
+ size_t event_data_bytes,
+ size_t event_entry_count,
+ size_t offset, void *end)
+{
+ ssize_t ev_len;
+ void *ev_end, *calc_ev_end;
+
+ if (offset >= event_data_bytes)
+ return -1;
+
+ if (event_idx >= event_entry_count) {
+ pr_devel("catalog event data has %zu bytes of padding after last event\n",
+ event_data_bytes - offset);
+ return -1;
+ }
+
+ if (!event_fixed_portion_is_within(event, end)) {
+ pr_warn("event %zu fixed portion is not within range\n",
+ event_idx);
+ return -1;
+ }
+
+ ev_len = be16_to_cpu(event->length);
+
+ if (ev_len % 16)
+ pr_info("event %zu has length %zu not divisible by 16: event=%pK\n",
+ event_idx, ev_len, event);
+
+ ev_end = (__u8 *)event + ev_len;
+ if (ev_end > end) {
+ pr_warn("event %zu has .length=%zu, ends after buffer end: ev_end=%pK > end=%pK, offset=%zu\n",
+ event_idx, ev_len, ev_end, end,
+ offset);
+ return -1;
+ }
+
+ calc_ev_end = event_end(event, end);
+ if (!calc_ev_end) {
+ pr_warn("event %zu has a calculated length which exceeds buffer length %zu: event=%pK end=%pK, offset=%zu\n",
+ event_idx, event_data_bytes, event, end,
+ offset);
+ return -1;
+ }
+
+ if (calc_ev_end > ev_end) {
+ pr_warn("event %zu exceeds it's own length: event=%pK, end=%pK, offset=%zu, calc_ev_end=%pK\n",
+ event_idx, event, ev_end, offset, calc_ev_end);
+ return -1;
+ }
+
+ return ev_len;
+}
+
+#define MAX_4K (SIZE_MAX / 4096)
+
+static void create_events_from_catalog(struct attribute ***events_,
+ struct attribute ***event_descs_,
+ struct attribute ***event_long_descs_)
+{
+ unsigned long hret;
+ size_t catalog_len, catalog_page_len, event_entry_count,
+ event_data_len, event_data_offs,
+ event_data_bytes, junk_events, event_idx, event_attr_ct, i,
+ attr_max, event_idx_last, desc_ct, long_desc_ct;
+ ssize_t ct, ev_len;
+ uint32_t catalog_version_num;
+ struct attribute **events, **event_descs, **event_long_descs;
+ struct hv_24x7_catalog_page_0 *page_0 =
+ kmem_cache_alloc(hv_page_cache, GFP_KERNEL);
+ void *page = page_0;
+ void *event_data, *end;
+ struct hv_24x7_event_data *event;
+ struct rb_root ev_uniq = RB_ROOT;
+
+ if (!page)
+ goto e_out;
+
+ hret = h_get_24x7_catalog_page(page, 0, 0);
+ if (hret)
+ goto e_free;
+
+ catalog_version_num = be64_to_cpu(page_0->version);
+ catalog_page_len = be32_to_cpu(page_0->length);
+
+ if (MAX_4K < catalog_page_len) {
+ pr_err("invalid page count: %zu\n", catalog_page_len);
+ goto e_free;
+ }
+
+ catalog_len = catalog_page_len * 4096;
+
+ event_entry_count = be16_to_cpu(page_0->event_entry_count);
+ event_data_offs = be16_to_cpu(page_0->event_data_offs);
+ event_data_len = be16_to_cpu(page_0->event_data_len);
+
+ pr_devel("cv %zu cl %zu eec %zu edo %zu edl %zu\n",
+ (size_t)catalog_version_num, catalog_len,
+ event_entry_count, event_data_offs, event_data_len);
+
+ if ((MAX_4K < event_data_len)
+ || (MAX_4K < event_data_offs)
+ || (MAX_4K - event_data_offs < event_data_len)) {
+ pr_err("invalid event data offs %zu and/or len %zu\n",
+ event_data_offs, event_data_len);
+ goto e_free;
+ }
+
+ if ((event_data_offs + event_data_len) > catalog_page_len) {
+ pr_err("event data %zu-%zu does not fit inside catalog 0-%zu\n",
+ event_data_offs,
+ event_data_offs + event_data_len,
+ catalog_page_len);
+ goto e_free;
+ }
+
+ if (SIZE_MAX / MAX_EVENTS_PER_EVENT_DATA - 1 < event_entry_count) {
+ pr_err("event_entry_count %zu is invalid\n",
+ event_entry_count);
+ goto e_free;
+ }
+
+ event_data_bytes = event_data_len * 4096;
+
+ /*
+ * event data can span several pages, events can cross between these
+ * pages. Use vmalloc to make this easier.
+ */
+ event_data = vmalloc(event_data_bytes);
+ if (!event_data) {
+ pr_err("could not allocate event data\n");
+ goto e_free;
+ }
+
+ end = event_data + event_data_bytes;
+
+ /*
+ * using vmalloc_to_phys() like this only works if PAGE_SIZE is
+ * divisible by 4096
+ */
+ BUILD_BUG_ON(PAGE_SIZE % 4096);
+
+ for (i = 0; i < event_data_len; i++) {
+ hret = h_get_24x7_catalog_page_(
+ vmalloc_to_phys(event_data + i * 4096),
+ catalog_version_num,
+ i + event_data_offs);
+ if (hret) {
+ pr_err("failed to get event data in page %zu\n",
+ i + event_data_offs);
+ goto e_event_data;
+ }
+ }
+
+ /*
+ * scan the catalog to determine the number of attributes we need, and
+ * verify it at the same time.
+ */
+ for (junk_events = 0, event = event_data, event_idx = 0, attr_max = 0;
+ ;
+ event_idx++, event = (void *)event + ev_len) {
+ size_t offset = (void *)event - (void *)event_data;
+ char *name;
+ int nl;
+
+ ev_len = catalog_event_len_validate(event, event_idx,
+ event_data_bytes,
+ event_entry_count,
+ offset, end);
+ if (ev_len < 0)
+ break;
+
+ name = event_name(event, &nl);
+
+ if (event->event_group_record_len == 0) {
+ pr_devel("invalid event %zu (%.*s): group_record_len == 0, skipping\n",
+ event_idx, nl, name);
+ junk_events++;
+ continue;
+ }
+
+ if (!catalog_entry_domain_is_valid(event->domain)) {
+ pr_info("event %zu (%.*s) has invalid domain %d\n",
+ event_idx, nl, name, event->domain);
+ junk_events++;
+ continue;
+ }
+
+ attr_max += event_to_attr_ct(event);
+ }
+
+ event_idx_last = event_idx;
+ if (event_idx_last != event_entry_count)
+ pr_warn("event buffer ended before listed # of events were parsed (got %zu, wanted %zu, junk %zu)\n",
+ event_idx_last, event_entry_count, junk_events);
+
+ events = kmalloc_array(attr_max + 1, sizeof(*events), GFP_KERNEL);
+ if (!events)
+ goto e_event_data;
+
+ event_descs = kmalloc_array(event_idx + 1, sizeof(*event_descs),
+ GFP_KERNEL);
+ if (!event_descs)
+ goto e_event_attrs;
+
+ event_long_descs = kmalloc_array(event_idx + 1,
+ sizeof(*event_long_descs), GFP_KERNEL);
+ if (!event_long_descs)
+ goto e_event_descs;
+
+ /* Iterate over the catalog filling in the attribute vector */
+ for (junk_events = 0, event_attr_ct = 0, desc_ct = 0, long_desc_ct = 0,
+ event = event_data, event_idx = 0;
+ event_idx < event_idx_last;
+ event_idx++, ev_len = be16_to_cpu(event->length),
+ event = (void *)event + ev_len) {
+ char *name;
+ int nl;
+ int nonce;
+ /*
+ * these are the only "bad" events that are intermixed and that
+ * we can ignore without issue. make sure to skip them here
+ */
+ if (event->event_group_record_len == 0)
+ continue;
+ if (!catalog_entry_domain_is_valid(event->domain))
+ continue;
+
+ name = event_name(event, &nl);
+ nonce = event_uniq_add(&ev_uniq, name, nl, event->domain);
+ ct = event_data_to_attrs(event_idx, events + event_attr_ct,
+ event, nonce);
+ if (ct <= 0) {
+ pr_warn("event %zu (%.*s) creation failure, skipping\n",
+ event_idx, nl, name);
+ junk_events++;
+ } else {
+ event_attr_ct += ct;
+ event_descs[desc_ct] = event_to_desc_attr(event, nonce);
+ if (event_descs[desc_ct])
+ desc_ct++;
+ event_long_descs[long_desc_ct] =
+ event_to_long_desc_attr(event, nonce);
+ if (event_long_descs[long_desc_ct])
+ long_desc_ct++;
+ }
+ }
+
+ pr_info("read %zu catalog entries, created %zu event attrs (%zu failures), %zu descs\n",
+ event_idx, event_attr_ct, junk_events, desc_ct);
+
+ events[event_attr_ct] = NULL;
+ event_descs[desc_ct] = NULL;
+ event_long_descs[long_desc_ct] = NULL;
+
+ event_uniq_destroy(&ev_uniq);
+ vfree(event_data);
+ kmem_cache_free(hv_page_cache, page);
+
+ *events_ = events;
+ *event_descs_ = event_descs;
+ *event_long_descs_ = event_long_descs;
+ return;
+
+e_event_descs:
+ kfree(event_descs);
+e_event_attrs:
+ kfree(events);
+e_event_data:
+ vfree(event_data);
+e_free:
+ kmem_cache_free(hv_page_cache, page);
+e_out:
+ *events_ = NULL;
+ *event_descs_ = NULL;
+ *event_long_descs_ = NULL;
+}
+
static ssize_t catalog_read(struct file *filp, struct kobject *kobj,
struct bin_attribute *bin_attr, char *buf,
loff_t offset, size_t count)
@@ -178,7 +948,7 @@ static ssize_t _name##_show(struct device *dev, \
} \
ret = sprintf(buf, _fmt, _expr); \
e_free: \
- kfree(page); \
+ kmem_cache_free(hv_page_cache, page); \
return ret; \
} \
static DEVICE_ATTR_RO(_name)
@@ -208,16 +978,13 @@ static struct attribute_group if_group = {
static const struct attribute_group *attr_groups[] = {
&format_group,
+ &event_group,
+ &event_desc_group,
+ &event_long_desc_group,
&if_group,
NULL,
};
-static bool is_physical_domain(int domain)
-{
- return domain == HV_24X7_PERF_DOMAIN_PHYSICAL_CHIP ||
- domain == HV_24X7_PERF_DOMAIN_PHYSICAL_CORE;
-}
-
static unsigned long single_24x7_request(u8 domain, u32 offset, u16 ix,
u16 lpar, u64 *res,
bool success_expected)
@@ -363,7 +1130,7 @@ static int h_24x7_event_init(struct perf_event *event)
return -EIO;
}
- /* PHYSICAL domains & other lpars require extra capabilities */
+ /* Physical domains & other lpars require extra capabilities */
if (!caps.collect_privileged && (is_physical_domain(domain) ||
(event_get_lpar(event) != event_get_lpar_max()))) {
pr_devel("hv permisions disallow: is_physical_domain:%d, lpar=0x%llx\n",
@@ -465,6 +1232,10 @@ static int hv_24x7_init(void)
/* sampling not supported */
h_24x7_pmu.capabilities |= PERF_PMU_CAP_NO_INTERRUPT;
+ create_events_from_catalog(&event_group.attrs,
+ &event_desc_group.attrs,
+ &event_long_desc_group.attrs);
+
r = perf_pmu_register(&h_24x7_pmu, h_24x7_pmu.name, -1);
if (r)
return r;
diff --git a/arch/powerpc/perf/hv-24x7.h b/arch/powerpc/perf/hv-24x7.h
index 720ebce..69cd4e6 100644
--- a/arch/powerpc/perf/hv-24x7.h
+++ b/arch/powerpc/perf/hv-24x7.h
@@ -3,14 +3,14 @@
#include <linux/types.h>
+enum hv_perf_domains {
+#define DOMAIN(n, v, x, c) HV_PERF_DOMAIN_##n = v,
+#include "hv-24x7-domains.h"
+#undef DOMAIN
+};
+
struct hv_24x7_request {
/* PHYSICAL domains require enabling via phyp/hmc. */
-#define HV_24X7_PERF_DOMAIN_PHYSICAL_CHIP 0x01
-#define HV_24X7_PERF_DOMAIN_PHYSICAL_CORE 0x02
-#define HV_24X7_PERF_DOMAIN_VIRTUAL_PROCESSOR_HOME_CORE 0x03
-#define HV_24X7_PERF_DOMAIN_VIRTUAL_PROCESSOR_HOME_CHIP 0x04
-#define HV_24X7_PERF_DOMAIN_VIRTUAL_PROCESSOR_HOME_NODE 0x05
-#define HV_24X7_PERF_DOMAIN_VIRTUAL_PROCESSOR_REMOTE_NODE 0x06
__u8 performance_domain;
__u8 reserved[0x1];
--
1.7.9.5
^ permalink raw reply related
* [PATCH v3 03/10] perf: provide sysfs_show for struct perf_pmu_events_attr
From: Sukadev Bhattiprolu @ 2014-09-17 18:51 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, ak, Jiri Olsa, peterz, eranian,
Paul Mackerras
Cc: Michael Ellerman, linux-kernel, dev, linuxppc-dev,
Anshuman Khandual
In-Reply-To: <1410979923-7722-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
(struct perf_pmu_events_attr) is defined in include/linux/perf_event.h,
but the only "show" for it is in x86 and contains x86 specific stuff.
Make a generic one for those of us who are just using the event_str.
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
include/linux/perf_event.h | 3 +++
kernel/events/core.c | 8 ++++++++
2 files changed, 11 insertions(+)
diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index 707617a..cef4a56 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -882,6 +882,9 @@ struct perf_pmu_events_attr {
const char *event_str;
};
+ssize_t perf_event_sysfs_show(struct device *dev, struct device_attribute *attr,
+ char *page);
+
#define PMU_EVENT_ATTR(_name, _var, _id, _show) \
static struct perf_pmu_events_attr _var = { \
.attr = __ATTR(_name, 0444, _show, NULL), \
diff --git a/kernel/events/core.c b/kernel/events/core.c
index f9c1ed0..cf1a423 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -8100,6 +8100,14 @@ void __init perf_event_init(void)
!= 1024);
}
+ssize_t perf_event_sysfs_show(struct device *dev, struct device_attribute *attr,
+ char *page)
+{
+ struct perf_pmu_events_attr *pmu_attr =
+ container_of(attr, struct perf_pmu_events_attr, attr);
+ return sprintf(page, "%s\n", pmu_attr->event_str);
+}
+
static int __init perf_event_sysfs_init(void)
{
struct pmu *pmu;
--
1.7.9.5
^ permalink raw reply related
* [PATCH v3 01/10] tools/perf: support parsing parameterized events
From: Sukadev Bhattiprolu @ 2014-09-17 18:51 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, ak, Jiri Olsa, peterz, eranian,
Paul Mackerras
Cc: Michael Ellerman, linux-kernel, dev, linuxppc-dev,
Anshuman Khandual
In-Reply-To: <1410979923-7722-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
Enable event specification like:
pmu/event_name,param1=0x1,param2=0x4/
Assuming that
/sys/bus/event_source/devices/pmu/events/event_name
Contains something like
param2=foo,bar=1,param1=baz
Changelog[v6]:
[Jiri Olsa] If the sysfs event file specifies 'param=val', make the
usage 'hv_24x7/event,param=123/' rather than 'hv_24x7/event,val=123/'.
[Sukadev Bhattiprolu] If verbose option is set, print help message
if a required parameter is missing or invalid parameter is specified.
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
tools/perf/util/parse-events.h | 1 +
tools/perf/util/pmu.c | 64 +++++++++++++++++++++++++++++++++-------
2 files changed, 54 insertions(+), 11 deletions(-)
diff --git a/tools/perf/util/parse-events.h b/tools/perf/util/parse-events.h
index df094b4..9d7d2d5 100644
--- a/tools/perf/util/parse-events.h
+++ b/tools/perf/util/parse-events.h
@@ -59,6 +59,7 @@ struct parse_events_term {
int type_val;
int type_term;
struct list_head list;
+ bool used;
};
struct parse_events_evlist {
diff --git a/tools/perf/util/pmu.c b/tools/perf/util/pmu.c
index 9bf5827..8c7c4a1 100644
--- a/tools/perf/util/pmu.c
+++ b/tools/perf/util/pmu.c
@@ -504,30 +504,67 @@ static __u64 pmu_format_value(unsigned long *format, __u64 value)
}
/*
+ * Term is a string term, and might be a param-term. Try to look up it's value
+ * in the remaining terms.
+ * - We have a term like "base-or-format-term=param-term",
+ * - We need to find the value supplied for "param-term" (with param-term named
+ * in a config string) later on in the term list.
+ */
+static int pmu_resolve_param_term(struct parse_events_term *term,
+ struct list_head *head_terms,
+ __u64 *value)
+{
+ struct parse_events_term *t;
+
+ list_for_each_entry(t, head_terms, list) {
+ if (t->type_val == PARSE_EVENTS__TERM_TYPE_NUM) {
+ if (!strcmp(t->config, term->config)) {
+ t->used = true;
+ *value = t->val.num;
+ return 0;
+ }
+ }
+ }
+
+ if (verbose)
+ printf("Required parameter '%s' not specified\n", term->config);
+
+ return -1;
+}
+
+/*
* Setup one of config[12] attr members based on the
* user input data - term parameter.
*/
static int pmu_config_term(struct list_head *formats,
struct perf_event_attr *attr,
- struct parse_events_term *term)
+ struct parse_events_term *term,
+ struct list_head *head_terms)
{
struct perf_pmu_format *format;
__u64 *vp;
+ __u64 val;
+
+ /*
+ * If this is a parameter we've already used for parameterized-eval,
+ * skip it in normal eval.
+ */
+ if (term->used)
+ return 0;
/*
- * Support only for hardcoded and numnerial terms.
* Hardcoded terms should be already in, so nothing
* to be done for them.
*/
if (parse_events__is_hardcoded_term(term))
return 0;
- if (term->type_val != PARSE_EVENTS__TERM_TYPE_NUM)
- return -EINVAL;
-
format = pmu_find_format(formats, term->config);
- if (!format)
+ if (!format) {
+ if (verbose)
+ printf("Invalid event/parameter '%s'\n", term->config);
return -EINVAL;
+ }
switch (format->value) {
case PERF_PMU_FORMAT_VALUE_CONFIG:
@@ -544,11 +581,16 @@ static int pmu_config_term(struct list_head *formats,
}
/*
- * XXX If we ever decide to go with string values for
- * non-hardcoded terms, here's the place to translate
- * them into value.
+ * Either directly use a numeric term, or try to translate string terms
+ * using event parameters.
*/
- *vp |= pmu_format_value(format->bits, term->val.num);
+ if (term->type_val == PARSE_EVENTS__TERM_TYPE_NUM)
+ val = term->val.num;
+ else
+ if (pmu_resolve_param_term(term, head_terms, &val))
+ return -EINVAL;
+
+ *vp |= pmu_format_value(format->bits, val);
return 0;
}
@@ -559,7 +601,7 @@ int perf_pmu__config_terms(struct list_head *formats,
struct parse_events_term *term;
list_for_each_entry(term, head_terms, list)
- if (pmu_config_term(formats, attr, term))
+ if (pmu_config_term(formats, attr, term, head_terms))
return -EINVAL;
return 0;
--
1.7.9.5
^ permalink raw reply related
* [PATCH v3 02/10] tools/perf: extend format_alias() to include event parameters
From: Sukadev Bhattiprolu @ 2014-09-17 18:51 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, ak, Jiri Olsa, peterz, eranian,
Paul Mackerras
Cc: Michael Ellerman, linux-kernel, dev, linuxppc-dev,
Anshuman Khandual
In-Reply-To: <1410979923-7722-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
This causes `perf list pmu` to show parameters for parameterized events
like follows:
pmu/event_name,param1=?,param2=?/ [Kernel PMU event]
An example:
hv_gpci/dispatch_timebase_by_processor_processor_time_in_timebase_cycles,phys_processor_idx=?/ [Kernel PMU event]
Changelog[v6]
[Jir Olsa] If the parameter for an event in sysfs is 'param=val',
have perf-list show the event as 'param=?' rather than 'val=?'.
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
---
tools/perf/util/pmu.c | 26 +++++++++++++++++++++++++-
1 file changed, 25 insertions(+), 1 deletion(-)
diff --git a/tools/perf/util/pmu.c b/tools/perf/util/pmu.c
index 8c7c4a1..0756917 100644
--- a/tools/perf/util/pmu.c
+++ b/tools/perf/util/pmu.c
@@ -748,10 +748,33 @@ void perf_pmu__set_format(unsigned long *bits, long from, long to)
set_bit(b, bits);
}
+static int sub_non_neg(int a, int b)
+{
+ if (b > a)
+ return 0;
+ return a - b;
+}
+
static char *format_alias(char *buf, int len, struct perf_pmu *pmu,
struct perf_pmu_alias *alias)
{
- snprintf(buf, len, "%s/%s/", pmu->name, alias->name);
+ struct parse_events_term *term;
+ int used = snprintf(buf, len, "%s/%s", pmu->name, alias->name);
+
+ list_for_each_entry(term, &alias->terms, list)
+ if (term->type_val == PARSE_EVENTS__TERM_TYPE_STR)
+ used += snprintf(buf + used, sub_non_neg(len, used),
+ ",%s=?", term->config);
+
+ if (sub_non_neg(len, used) > 0) {
+ buf[used] = '/';
+ used++;
+ }
+ if (sub_non_neg(len, used) > 0) {
+ buf[used] = '\0';
+ used++;
+ } else
+ buf[len - 1] = '\0';
return buf;
}
@@ -802,6 +825,7 @@ void print_pmu_events(const char *event_glob, bool name_only)
if (is_cpu && !name_only)
aliases[j] = format_alias_or(buf, sizeof(buf),
pmu, alias);
+
aliases[j] = strdup(aliases[j]);
j++;
}
--
1.7.9.5
^ permalink raw reply related
* [PATCH v3 00/10] Add support for parameterized events from sysfs
From: Sukadev Bhattiprolu @ 2014-09-17 18:51 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, ak, Jiri Olsa, peterz, eranian,
Paul Mackerras
Cc: Michael Ellerman, linux-kernel, dev, linuxppc-dev,
Anshuman Khandual
What this patchset does:
- the first patch (override sysfs in tools/perf via SYSFS_PATH) was sent out
previously, but needed a resend anyhow. Having it is useful for testing the
later changes to tools/perf.
- the second patch is a bugfix to the powerpc hv-24x7 code which was
previously sent out, which is a good idea to have when testing these patches
on POWER8 hardware.
- document perf sysfs and the changes to add parameterized events
- semi-notably: removes the growing list of specific POWER cpu events and
begins documenting them generically, much like the docs for
/sys/modules/MODULENAME do for modules.
- tools/perf changes to support parameterized events
- export some parameterized events from the powerpc pmus hv_24x7 and hv_gpci
Description of "event parameters" from the documentation patch:
Event parameters are a basic way for partial events to be specified in
sysfs with per-event names given to the fields that need to be filled in
when using a particular event.
It is intended for supporting cases where the single 'cpu' parameter is
insufficient. For example, POWER 8 has events for physical
sockets/cores/cpus that are accessible from with virtual machines. To
keep using the single 'cpu' parameter we'd need to perform a mapping
between Linux's cpus and the physical machine's cpus (in this case
Linux is running under a hypervisor). This isn't possible because
bindings between our cpus and physical cpus may not be fixed, and we
probably won't have a "cpu" on each physical cpu.
Description of the sysfs contents when events are parameterized (copied from an
included patch):
Examples:
domain=0x1,offset=0x8,starting_index=phys_cpu
In the case of the last example, a value replacing "phys_cpu"
would need to be provided by the user selecting the particular
event. This is refered to as "event parameterization". All
non-numerical values indicate an event parameter.
Notes on how perf-list displays parameterized events (and how to use them,
again culled from an included patch):
PARAMETERIZED EVENTS
--------------------
Some pmu events listed by 'perf-list' will be displayed with '?' in
them. For example:
hv_gpci/dtbp_ptitc,phys_processor_idx=?/
This means that when provided as an event, a value for
phys_processor_idx must also be supplied. For example:
perf stat -e 'hv_gpci/dtbp_ptitc,phys_processor_idx=0x2/' ...
Changelog[v3]
- [Jiri Olsa] Changed the event parameters are specified. If
event file specifes 'param=val' make the usage 'param=123'
rather than 'val=123'. (patch 1,2/10)
- Shortened event names using "PHYS" and "VCPU" (patch 4/10)
- Print help message if invalid parameter is specified or required
parameter is missing.
- Moved 3 patches that are unrelated to parametrized events into
a separate patchset.
- Reordered patches so code changes come first.
Changelog[v2]
- [Joe Perches, David Laight] Use beNN_to_cpu() instead of guessing
the size from type.
- Use kmem_cache_free() to free page allocated with kmem_cache_alloc().
- Rebase to recent kernel
*** BLURB HERE ***
Cody P Schafer (10):
tools/perf: support parsing parameterized events
tools/perf: extend format_alias() to include event parameters
perf: provide sysfs_show for struct perf_pmu_events_attr
powerpc/perf/hv-24x7: parse catalog and populate sysfs with events
perf: add PMU_EVENT_ATTR_STRING() helper
powerpc/perf/{hv-gpci,hv-common}: generate requests with counters
annotated
powerpc/perf/hv-gpci: add the remaining gpci requests
perf Documentation: add event parameters
tools/perf: Document parameterized and symbolic events
powerpc/perf/hv-24x7: Document sysfs event description entries
.../testing/sysfs-bus-event_source-devices-events | 6 +
.../testing/sysfs-bus-event_source-devices-hv_24x7 | 22 +
arch/powerpc/perf/hv-24x7-catalog.h | 25 +
arch/powerpc/perf/hv-24x7-domains.h | 28 +
arch/powerpc/perf/hv-24x7.c | 787 +++++++++++++++++++-
arch/powerpc/perf/hv-24x7.h | 12 +-
arch/powerpc/perf/hv-common.c | 10 +-
arch/powerpc/perf/hv-gpci-requests.h | 262 +++++++
arch/powerpc/perf/hv-gpci.c | 8 +
arch/powerpc/perf/hv-gpci.h | 37 +-
arch/powerpc/perf/req-gen/_begin.h | 13 +
arch/powerpc/perf/req-gen/_clear.h | 5 +
arch/powerpc/perf/req-gen/_end.h | 4 +
arch/powerpc/perf/req-gen/_request-begin.h | 15 +
arch/powerpc/perf/req-gen/_request-end.h | 8 +
arch/powerpc/perf/req-gen/perf.h | 155 ++++
include/linux/perf_event.h | 10 +
kernel/events/core.c | 8 +
tools/perf/Documentation/perf-list.txt | 13 +
tools/perf/Documentation/perf-record.txt | 12 +
tools/perf/Documentation/perf-stat.txt | 20 +-
tools/perf/util/parse-events.h | 1 +
tools/perf/util/pmu.c | 90 ++-
23 files changed, 1491 insertions(+), 60 deletions(-)
create mode 100644 arch/powerpc/perf/hv-24x7-domains.h
create mode 100644 arch/powerpc/perf/hv-gpci-requests.h
create mode 100644 arch/powerpc/perf/req-gen/_begin.h
create mode 100644 arch/powerpc/perf/req-gen/_clear.h
create mode 100644 arch/powerpc/perf/req-gen/_end.h
create mode 100644 arch/powerpc/perf/req-gen/_request-begin.h
create mode 100644 arch/powerpc/perf/req-gen/_request-end.h
create mode 100644 arch/powerpc/perf/req-gen/perf.h
--
1.7.9.5
^ permalink raw reply
* [PATCH 3/4] perf Documentation: sysfs events/ interfaces
From: Sukadev Bhattiprolu @ 2014-09-17 18:33 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, ak, Jiri Olsa, peterz, eranian,
Paul Mackerras
Cc: linuxppc-dev, dev, linux-kernel, Michael Ellerman,
Anshuman Khandual
In-Reply-To: <1410978787-7375-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
Add documentation for the <event>, <event>.scale, and <event>.unit
files in sysfs.
<event>.scale and <event>.unit were undocumented.
<event> was previously documented only for specific powerpc pmu events.
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
.../testing/sysfs-bus-event_source-devices-events | 60 ++++++++++++++++++++
1 file changed, 60 insertions(+)
diff --git a/Documentation/ABI/testing/sysfs-bus-event_source-devices-events b/Documentation/ABI/testing/sysfs-bus-event_source-devices-events
index 7b40a3c..a5226f0 100644
--- a/Documentation/ABI/testing/sysfs-bus-event_source-devices-events
+++ b/Documentation/ABI/testing/sysfs-bus-event_source-devices-events
@@ -599,3 +599,63 @@ Description: POWER-systems specific performance monitoring events
Further, multiple terms like 'event=0xNNNN' can be specified
and separated with comma. All available terms are defined in
the /sys/bus/event_source/devices/<dev>/format file.
+
+What: /sys/bus/event_source/devices/<pmu>/events/<event>
+Date: 2014/02/24
+Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
+Description: Per-pmu performance monitoring events specific to the running system
+
+ Each file (except for some of those with a '.' in them, '.unit'
+ and '.scale') in the 'events' directory describes a single
+ performance monitoring event supported by the <pmu>. The name
+ of the file is the name of the event.
+
+ File contents:
+
+ <term>[=<value>][,<term>[=<value>]]...
+
+ Where <term> is one of the terms listed under
+ /sys/bus/event_source/devices/<pmu>/format/ and <value> is
+ a number is base-16 format with a '0x' prefix (lowercase only).
+ If a <term> is specified alone (without an assigned value), it
+ is implied that 0x1 is assigned to that <term>.
+
+ Examples (each of these lines would be in a seperate file):
+
+ event=0x2abc
+ event=0x423,inv,cmask=0x3
+ domain=0x1,offset=0x8,starting_index=0xffff
+
+ Each of the assignments indicates a value to be assigned to a
+ particular set of bits (as defined by the format file
+ corresponding to the <term>) in the perf_event structure passed
+ to the perf_open syscall.
+
+What: /sys/bus/event_source/devices/<pmu>/events/<event>.unit
+Date: 2014/02/24
+Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
+Description: Perf event units
+
+ A string specifying the English plural numerical unit that <event>
+ (once multiplied by <event>.scale) represents.
+
+ Example:
+
+ Joules
+
+What: /sys/bus/event_source/devices/<pmu>/events/<event>.scale
+Date: 2014/02/24
+Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
+Description: Perf event scaling factors
+
+ A string representing a floating point value expressed in
+ scientific notation to be multiplied by the event count
+ recieved from the kernel to match the unit specified in the
+ <event>.unit file.
+
+ Example:
+
+ 2.3283064365386962890625e-10
+
+ This is provided to avoid performing floating point arithmetic
+ in the kernel.
--
1.7.9.5
^ permalink raw reply related
* [PATCH 4/4] perf Documentation: remove duplicated docs for powerpc cpu specific events
From: Sukadev Bhattiprolu @ 2014-09-17 18:33 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, ak, Jiri Olsa, peterz, eranian,
Paul Mackerras
Cc: linuxppc-dev, dev, linux-kernel, Michael Ellerman,
Anshuman Khandual
In-Reply-To: <1410978787-7375-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
Listing specific events doesn't actually help us at all here because:
- these events actually vary between different ppc processors, they
aren't garunteed to be present.
- the documentation of the (generic) file contents is now superceded by the
docs for arbitrary event file contents.
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
.../testing/sysfs-bus-event_source-devices-events | 573 --------------------
1 file changed, 573 deletions(-)
diff --git a/Documentation/ABI/testing/sysfs-bus-event_source-devices-events b/Documentation/ABI/testing/sysfs-bus-event_source-devices-events
index a5226f0..20979f8 100644
--- a/Documentation/ABI/testing/sysfs-bus-event_source-devices-events
+++ b/Documentation/ABI/testing/sysfs-bus-event_source-devices-events
@@ -27,579 +27,6 @@ Description: Generic performance monitoring events
"basename".
-What: /sys/devices/cpu/events/PM_1PLUS_PPC_CMPL
- /sys/devices/cpu/events/PM_BRU_FIN
- /sys/devices/cpu/events/PM_BR_MPRED
- /sys/devices/cpu/events/PM_CMPLU_STALL
- /sys/devices/cpu/events/PM_CMPLU_STALL_BRU
- /sys/devices/cpu/events/PM_CMPLU_STALL_DCACHE_MISS
- /sys/devices/cpu/events/PM_CMPLU_STALL_DFU
- /sys/devices/cpu/events/PM_CMPLU_STALL_DIV
- /sys/devices/cpu/events/PM_CMPLU_STALL_ERAT_MISS
- /sys/devices/cpu/events/PM_CMPLU_STALL_FXU
- /sys/devices/cpu/events/PM_CMPLU_STALL_IFU
- /sys/devices/cpu/events/PM_CMPLU_STALL_LSU
- /sys/devices/cpu/events/PM_CMPLU_STALL_REJECT
- /sys/devices/cpu/events/PM_CMPLU_STALL_SCALAR
- /sys/devices/cpu/events/PM_CMPLU_STALL_SCALAR_LONG
- /sys/devices/cpu/events/PM_CMPLU_STALL_STORE
- /sys/devices/cpu/events/PM_CMPLU_STALL_THRD
- /sys/devices/cpu/events/PM_CMPLU_STALL_VECTOR
- /sys/devices/cpu/events/PM_CMPLU_STALL_VECTOR_LONG
- /sys/devices/cpu/events/PM_CYC
- /sys/devices/cpu/events/PM_GCT_NOSLOT_BR_MPRED
- /sys/devices/cpu/events/PM_GCT_NOSLOT_BR_MPRED_IC_MISS
- /sys/devices/cpu/events/PM_GCT_NOSLOT_CYC
- /sys/devices/cpu/events/PM_GCT_NOSLOT_IC_MISS
- /sys/devices/cpu/events/PM_GRP_CMPL
- /sys/devices/cpu/events/PM_INST_CMPL
- /sys/devices/cpu/events/PM_LD_MISS_L1
- /sys/devices/cpu/events/PM_LD_REF_L1
- /sys/devices/cpu/events/PM_RUN_CYC
- /sys/devices/cpu/events/PM_RUN_INST_CMPL
- /sys/devices/cpu/events/PM_IC_DEMAND_L2_BR_ALL
- /sys/devices/cpu/events/PM_GCT_UTIL_7_TO_10_SLOTS
- /sys/devices/cpu/events/PM_PMC2_SAVED
- /sys/devices/cpu/events/PM_VSU0_16FLOP
- /sys/devices/cpu/events/PM_MRK_LSU_DERAT_MISS
- /sys/devices/cpu/events/PM_MRK_ST_CMPL
- /sys/devices/cpu/events/PM_NEST_PAIR3_ADD
- /sys/devices/cpu/events/PM_L2_ST_DISP
- /sys/devices/cpu/events/PM_L2_CASTOUT_MOD
- /sys/devices/cpu/events/PM_ISEG
- /sys/devices/cpu/events/PM_MRK_INST_TIMEO
- /sys/devices/cpu/events/PM_L2_RCST_DISP_FAIL_ADDR
- /sys/devices/cpu/events/PM_LSU1_DC_PREF_STREAM_CONFIRM
- /sys/devices/cpu/events/PM_IERAT_WR_64K
- /sys/devices/cpu/events/PM_MRK_DTLB_MISS_16M
- /sys/devices/cpu/events/PM_IERAT_MISS
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_LMEM
- /sys/devices/cpu/events/PM_FLOP
- /sys/devices/cpu/events/PM_THRD_PRIO_4_5_CYC
- /sys/devices/cpu/events/PM_BR_PRED_TA
- /sys/devices/cpu/events/PM_EXT_INT
- /sys/devices/cpu/events/PM_VSU_FSQRT_FDIV
- /sys/devices/cpu/events/PM_MRK_LD_MISS_EXPOSED_CYC
- /sys/devices/cpu/events/PM_LSU1_LDF
- /sys/devices/cpu/events/PM_IC_WRITE_ALL
- /sys/devices/cpu/events/PM_LSU0_SRQ_STFWD
- /sys/devices/cpu/events/PM_PTEG_FROM_RL2L3_MOD
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_L31_SHR
- /sys/devices/cpu/events/PM_DATA_FROM_L21_MOD
- /sys/devices/cpu/events/PM_VSU1_SCAL_DOUBLE_ISSUED
- /sys/devices/cpu/events/PM_VSU0_8FLOP
- /sys/devices/cpu/events/PM_POWER_EVENT1
- /sys/devices/cpu/events/PM_DISP_CLB_HELD_BAL
- /sys/devices/cpu/events/PM_VSU1_2FLOP
- /sys/devices/cpu/events/PM_LWSYNC_HELD
- /sys/devices/cpu/events/PM_PTEG_FROM_DL2L3_SHR
- /sys/devices/cpu/events/PM_INST_FROM_L21_MOD
- /sys/devices/cpu/events/PM_IERAT_XLATE_WR_16MPLUS
- /sys/devices/cpu/events/PM_IC_REQ_ALL
- /sys/devices/cpu/events/PM_DSLB_MISS
- /sys/devices/cpu/events/PM_L3_MISS
- /sys/devices/cpu/events/PM_LSU0_L1_PREF
- /sys/devices/cpu/events/PM_VSU_SCALAR_SINGLE_ISSUED
- /sys/devices/cpu/events/PM_LSU1_DC_PREF_STREAM_CONFIRM_STRIDE
- /sys/devices/cpu/events/PM_L2_INST
- /sys/devices/cpu/events/PM_VSU0_FRSP
- /sys/devices/cpu/events/PM_FLUSH_DISP
- /sys/devices/cpu/events/PM_PTEG_FROM_L2MISS
- /sys/devices/cpu/events/PM_VSU1_DQ_ISSUED
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_DMEM
- /sys/devices/cpu/events/PM_LSU_FLUSH_ULD
- /sys/devices/cpu/events/PM_PTEG_FROM_LMEM
- /sys/devices/cpu/events/PM_MRK_DERAT_MISS_16M
- /sys/devices/cpu/events/PM_THRD_ALL_RUN_CYC
- /sys/devices/cpu/events/PM_MEM0_PREFETCH_DISP
- /sys/devices/cpu/events/PM_MRK_STALL_CMPLU_CYC_COUNT
- /sys/devices/cpu/events/PM_DATA_FROM_DL2L3_MOD
- /sys/devices/cpu/events/PM_VSU_FRSP
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_L21_MOD
- /sys/devices/cpu/events/PM_PMC1_OVERFLOW
- /sys/devices/cpu/events/PM_VSU0_SINGLE
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_L3MISS
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_L31_SHR
- /sys/devices/cpu/events/PM_VSU0_VECTOR_SP_ISSUED
- /sys/devices/cpu/events/PM_VSU1_FEST
- /sys/devices/cpu/events/PM_MRK_INST_DISP
- /sys/devices/cpu/events/PM_VSU0_COMPLEX_ISSUED
- /sys/devices/cpu/events/PM_LSU1_FLUSH_UST
- /sys/devices/cpu/events/PM_FXU_IDLE
- /sys/devices/cpu/events/PM_LSU0_FLUSH_ULD
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_DL2L3_MOD
- /sys/devices/cpu/events/PM_LSU_LMQ_SRQ_EMPTY_ALL_CYC
- /sys/devices/cpu/events/PM_LSU1_REJECT_LMQ_FULL
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_L21_MOD
- /sys/devices/cpu/events/PM_INST_FROM_RL2L3_MOD
- /sys/devices/cpu/events/PM_SHL_CREATED
- /sys/devices/cpu/events/PM_L2_ST_HIT
- /sys/devices/cpu/events/PM_DATA_FROM_DMEM
- /sys/devices/cpu/events/PM_L3_LD_MISS
- /sys/devices/cpu/events/PM_FXU1_BUSY_FXU0_IDLE
- /sys/devices/cpu/events/PM_DISP_CLB_HELD_RES
- /sys/devices/cpu/events/PM_L2_SN_SX_I_DONE
- /sys/devices/cpu/events/PM_STCX_CMPL
- /sys/devices/cpu/events/PM_VSU0_2FLOP
- /sys/devices/cpu/events/PM_L3_PREF_MISS
- /sys/devices/cpu/events/PM_LSU_SRQ_SYNC_CYC
- /sys/devices/cpu/events/PM_LSU_REJECT_ERAT_MISS
- /sys/devices/cpu/events/PM_L1_ICACHE_MISS
- /sys/devices/cpu/events/PM_LSU1_FLUSH_SRQ
- /sys/devices/cpu/events/PM_LD_REF_L1_LSU0
- /sys/devices/cpu/events/PM_VSU0_FEST
- /sys/devices/cpu/events/PM_VSU_VECTOR_SINGLE_ISSUED
- /sys/devices/cpu/events/PM_FREQ_UP
- /sys/devices/cpu/events/PM_DATA_FROM_LMEM
- /sys/devices/cpu/events/PM_LSU1_LDX
- /sys/devices/cpu/events/PM_PMC3_OVERFLOW
- /sys/devices/cpu/events/PM_MRK_BR_MPRED
- /sys/devices/cpu/events/PM_SHL_MATCH
- /sys/devices/cpu/events/PM_MRK_BR_TAKEN
- /sys/devices/cpu/events/PM_ISLB_MISS
- /sys/devices/cpu/events/PM_DISP_HELD_THERMAL
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_RL2L3_SHR
- /sys/devices/cpu/events/PM_LSU1_SRQ_STFWD
- /sys/devices/cpu/events/PM_PTEG_FROM_DMEM
- /sys/devices/cpu/events/PM_VSU_2FLOP
- /sys/devices/cpu/events/PM_GCT_FULL_CYC
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_L3_CYC
- /sys/devices/cpu/events/PM_LSU_SRQ_S0_ALLOC
- /sys/devices/cpu/events/PM_MRK_DERAT_MISS_4K
- /sys/devices/cpu/events/PM_BR_MPRED_TA
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_L2MISS
- /sys/devices/cpu/events/PM_DPU_HELD_POWER
- /sys/devices/cpu/events/PM_MRK_VSU_FIN
- /sys/devices/cpu/events/PM_LSU_SRQ_S0_VALID
- /sys/devices/cpu/events/PM_GCT_EMPTY_CYC
- /sys/devices/cpu/events/PM_IOPS_DISP
- /sys/devices/cpu/events/PM_RUN_SPURR
- /sys/devices/cpu/events/PM_PTEG_FROM_L21_MOD
- /sys/devices/cpu/events/PM_VSU0_1FLOP
- /sys/devices/cpu/events/PM_SNOOP_TLBIE
- /sys/devices/cpu/events/PM_DATA_FROM_L3MISS
- /sys/devices/cpu/events/PM_VSU_SINGLE
- /sys/devices/cpu/events/PM_DTLB_MISS_16G
- /sys/devices/cpu/events/PM_FLUSH
- /sys/devices/cpu/events/PM_L2_LD_HIT
- /sys/devices/cpu/events/PM_NEST_PAIR2_AND
- /sys/devices/cpu/events/PM_VSU1_1FLOP
- /sys/devices/cpu/events/PM_IC_PREF_REQ
- /sys/devices/cpu/events/PM_L3_LD_HIT
- /sys/devices/cpu/events/PM_DISP_HELD
- /sys/devices/cpu/events/PM_L2_LD
- /sys/devices/cpu/events/PM_LSU_FLUSH_SRQ
- /sys/devices/cpu/events/PM_BC_PLUS_8_CONV
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_L31_MOD_CYC
- /sys/devices/cpu/events/PM_L2_RCST_BUSY_RC_FULL
- /sys/devices/cpu/events/PM_TB_BIT_TRANS
- /sys/devices/cpu/events/PM_THERMAL_MAX
- /sys/devices/cpu/events/PM_LSU1_FLUSH_ULD
- /sys/devices/cpu/events/PM_LSU1_REJECT_LHS
- /sys/devices/cpu/events/PM_LSU_LRQ_S0_ALLOC
- /sys/devices/cpu/events/PM_L3_CO_L31
- /sys/devices/cpu/events/PM_POWER_EVENT4
- /sys/devices/cpu/events/PM_DATA_FROM_L31_SHR
- /sys/devices/cpu/events/PM_BR_UNCOND
- /sys/devices/cpu/events/PM_LSU1_DC_PREF_STREAM_ALLOC
- /sys/devices/cpu/events/PM_PMC4_REWIND
- /sys/devices/cpu/events/PM_L2_RCLD_DISP
- /sys/devices/cpu/events/PM_THRD_PRIO_2_3_CYC
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_L2MISS
- /sys/devices/cpu/events/PM_IC_DEMAND_L2_BHT_REDIRECT
- /sys/devices/cpu/events/PM_DATA_FROM_L31_SHR
- /sys/devices/cpu/events/PM_IC_PREF_CANCEL_L2
- /sys/devices/cpu/events/PM_MRK_FIN_STALL_CYC_COUNT
- /sys/devices/cpu/events/PM_BR_PRED_CCACHE
- /sys/devices/cpu/events/PM_GCT_UTIL_1_TO_2_SLOTS
- /sys/devices/cpu/events/PM_MRK_ST_CMPL_INT
- /sys/devices/cpu/events/PM_LSU_TWO_TABLEWALK_CYC
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_L3MISS
- /sys/devices/cpu/events/PM_LSU_SET_MPRED
- /sys/devices/cpu/events/PM_FLUSH_DISP_TLBIE
- /sys/devices/cpu/events/PM_VSU1_FCONV
- /sys/devices/cpu/events/PM_DERAT_MISS_16G
- /sys/devices/cpu/events/PM_INST_FROM_LMEM
- /sys/devices/cpu/events/PM_IC_DEMAND_L2_BR_REDIRECT
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_L2
- /sys/devices/cpu/events/PM_PTEG_FROM_L2
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_L21_SHR_CYC
- /sys/devices/cpu/events/PM_MRK_DTLB_MISS_4K
- /sys/devices/cpu/events/PM_VSU0_FPSCR
- /sys/devices/cpu/events/PM_VSU1_VECT_DOUBLE_ISSUED
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_RL2L3_MOD
- /sys/devices/cpu/events/PM_MEM0_RQ_DISP
- /sys/devices/cpu/events/PM_L2_LD_MISS
- /sys/devices/cpu/events/PM_VMX_RESULT_SAT_1
- /sys/devices/cpu/events/PM_L1_PREF
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_LMEM_CYC
- /sys/devices/cpu/events/PM_GRP_IC_MISS_NONSPEC
- /sys/devices/cpu/events/PM_PB_NODE_PUMP
- /sys/devices/cpu/events/PM_SHL_MERGED
- /sys/devices/cpu/events/PM_NEST_PAIR1_ADD
- /sys/devices/cpu/events/PM_DATA_FROM_L3
- /sys/devices/cpu/events/PM_LSU_FLUSH
- /sys/devices/cpu/events/PM_LSU_SRQ_SYNC_COUNT
- /sys/devices/cpu/events/PM_PMC2_OVERFLOW
- /sys/devices/cpu/events/PM_LSU_LDF
- /sys/devices/cpu/events/PM_POWER_EVENT3
- /sys/devices/cpu/events/PM_DISP_WT
- /sys/devices/cpu/events/PM_IC_BANK_CONFLICT
- /sys/devices/cpu/events/PM_BR_MPRED_CR_TA
- /sys/devices/cpu/events/PM_L2_INST_MISS
- /sys/devices/cpu/events/PM_NEST_PAIR2_ADD
- /sys/devices/cpu/events/PM_MRK_LSU_FLUSH
- /sys/devices/cpu/events/PM_L2_LDST
- /sys/devices/cpu/events/PM_INST_FROM_L31_SHR
- /sys/devices/cpu/events/PM_VSU0_FIN
- /sys/devices/cpu/events/PM_VSU1_FCONV
- /sys/devices/cpu/events/PM_INST_FROM_RMEM
- /sys/devices/cpu/events/PM_DISP_CLB_HELD_TLBIE
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_DMEM_CYC
- /sys/devices/cpu/events/PM_BR_PRED_CR
- /sys/devices/cpu/events/PM_LSU_REJECT
- /sys/devices/cpu/events/PM_GCT_UTIL_3_TO_6_SLOTS
- /sys/devices/cpu/events/PM_CMPLU_STALL_END_GCT_NOSLOT
- /sys/devices/cpu/events/PM_LSU0_REJECT_LMQ_FULL
- /sys/devices/cpu/events/PM_VSU_FEST
- /sys/devices/cpu/events/PM_NEST_PAIR0_AND
- /sys/devices/cpu/events/PM_PTEG_FROM_L3
- /sys/devices/cpu/events/PM_POWER_EVENT2
- /sys/devices/cpu/events/PM_IC_PREF_CANCEL_PAGE
- /sys/devices/cpu/events/PM_VSU0_FSQRT_FDIV
- /sys/devices/cpu/events/PM_MRK_GRP_CMPL
- /sys/devices/cpu/events/PM_VSU0_SCAL_DOUBLE_ISSUED
- /sys/devices/cpu/events/PM_GRP_DISP
- /sys/devices/cpu/events/PM_LSU0_LDX
- /sys/devices/cpu/events/PM_DATA_FROM_L2
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_RL2L3_MOD
- /sys/devices/cpu/events/PM_VSU0_VECT_DOUBLE_ISSUED
- /sys/devices/cpu/events/PM_VSU1_2FLOP_DOUBLE
- /sys/devices/cpu/events/PM_THRD_PRIO_6_7_CYC
- /sys/devices/cpu/events/PM_BC_PLUS_8_RSLV_TAKEN
- /sys/devices/cpu/events/PM_BR_MPRED_CR
- /sys/devices/cpu/events/PM_L3_CO_MEM
- /sys/devices/cpu/events/PM_DATA_FROM_RL2L3_MOD
- /sys/devices/cpu/events/PM_LSU_SRQ_FULL_CYC
- /sys/devices/cpu/events/PM_TABLEWALK_CYC
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_RMEM
- /sys/devices/cpu/events/PM_LSU_SRQ_STFWD
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_RMEM
- /sys/devices/cpu/events/PM_FXU0_FIN
- /sys/devices/cpu/events/PM_LSU1_L1_SW_PREF
- /sys/devices/cpu/events/PM_PTEG_FROM_L31_MOD
- /sys/devices/cpu/events/PM_PMC5_OVERFLOW
- /sys/devices/cpu/events/PM_LD_REF_L1_LSU1
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_L21_SHR
- /sys/devices/cpu/events/PM_DATA_FROM_RMEM
- /sys/devices/cpu/events/PM_VSU0_SCAL_SINGLE_ISSUED
- /sys/devices/cpu/events/PM_BR_MPRED_LSTACK
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_RL2L3_MOD_CYC
- /sys/devices/cpu/events/PM_LSU0_FLUSH_UST
- /sys/devices/cpu/events/PM_LSU_NCST
- /sys/devices/cpu/events/PM_BR_TAKEN
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_LMEM
- /sys/devices/cpu/events/PM_DTLB_MISS_4K
- /sys/devices/cpu/events/PM_PMC4_SAVED
- /sys/devices/cpu/events/PM_VSU1_PERMUTE_ISSUED
- /sys/devices/cpu/events/PM_SLB_MISS
- /sys/devices/cpu/events/PM_LSU1_FLUSH_LRQ
- /sys/devices/cpu/events/PM_DTLB_MISS
- /sys/devices/cpu/events/PM_VSU1_FRSP
- /sys/devices/cpu/events/PM_VSU_VECTOR_DOUBLE_ISSUED
- /sys/devices/cpu/events/PM_L2_CASTOUT_SHR
- /sys/devices/cpu/events/PM_DATA_FROM_DL2L3_SHR
- /sys/devices/cpu/events/PM_VSU1_STF
- /sys/devices/cpu/events/PM_ST_FIN
- /sys/devices/cpu/events/PM_PTEG_FROM_L21_SHR
- /sys/devices/cpu/events/PM_L2_LOC_GUESS_WRONG
- /sys/devices/cpu/events/PM_MRK_STCX_FAIL
- /sys/devices/cpu/events/PM_LSU0_REJECT_LHS
- /sys/devices/cpu/events/PM_IC_PREF_CANCEL_HIT
- /sys/devices/cpu/events/PM_L3_PREF_BUSY
- /sys/devices/cpu/events/PM_MRK_BRU_FIN
- /sys/devices/cpu/events/PM_LSU1_NCLD
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_L31_MOD
- /sys/devices/cpu/events/PM_LSU_NCLD
- /sys/devices/cpu/events/PM_LSU_LDX
- /sys/devices/cpu/events/PM_L2_LOC_GUESS_CORRECT
- /sys/devices/cpu/events/PM_THRESH_TIMEO
- /sys/devices/cpu/events/PM_L3_PREF_ST
- /sys/devices/cpu/events/PM_DISP_CLB_HELD_SYNC
- /sys/devices/cpu/events/PM_VSU_SIMPLE_ISSUED
- /sys/devices/cpu/events/PM_VSU1_SINGLE
- /sys/devices/cpu/events/PM_DATA_TABLEWALK_CYC
- /sys/devices/cpu/events/PM_L2_RC_ST_DONE
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_L21_MOD
- /sys/devices/cpu/events/PM_LARX_LSU1
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_RMEM
- /sys/devices/cpu/events/PM_DISP_CLB_HELD
- /sys/devices/cpu/events/PM_DERAT_MISS_4K
- /sys/devices/cpu/events/PM_L2_RCLD_DISP_FAIL_ADDR
- /sys/devices/cpu/events/PM_SEG_EXCEPTION
- /sys/devices/cpu/events/PM_FLUSH_DISP_SB
- /sys/devices/cpu/events/PM_L2_DC_INV
- /sys/devices/cpu/events/PM_PTEG_FROM_DL2L3_MOD
- /sys/devices/cpu/events/PM_DSEG
- /sys/devices/cpu/events/PM_BR_PRED_LSTACK
- /sys/devices/cpu/events/PM_VSU0_STF
- /sys/devices/cpu/events/PM_LSU_FX_FIN
- /sys/devices/cpu/events/PM_DERAT_MISS_16M
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_DL2L3_MOD
- /sys/devices/cpu/events/PM_GCT_UTIL_11_PLUS_SLOTS
- /sys/devices/cpu/events/PM_INST_FROM_L3
- /sys/devices/cpu/events/PM_MRK_IFU_FIN
- /sys/devices/cpu/events/PM_ITLB_MISS
- /sys/devices/cpu/events/PM_VSU_STF
- /sys/devices/cpu/events/PM_LSU_FLUSH_UST
- /sys/devices/cpu/events/PM_L2_LDST_MISS
- /sys/devices/cpu/events/PM_FXU1_FIN
- /sys/devices/cpu/events/PM_SHL_DEALLOCATED
- /sys/devices/cpu/events/PM_L2_SN_M_WR_DONE
- /sys/devices/cpu/events/PM_LSU_REJECT_SET_MPRED
- /sys/devices/cpu/events/PM_L3_PREF_LD
- /sys/devices/cpu/events/PM_L2_SN_M_RD_DONE
- /sys/devices/cpu/events/PM_MRK_DERAT_MISS_16G
- /sys/devices/cpu/events/PM_VSU_FCONV
- /sys/devices/cpu/events/PM_ANY_THRD_RUN_CYC
- /sys/devices/cpu/events/PM_LSU_LMQ_FULL_CYC
- /sys/devices/cpu/events/PM_MRK_LSU_REJECT_LHS
- /sys/devices/cpu/events/PM_MRK_LD_MISS_L1_CYC
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_L2_CYC
- /sys/devices/cpu/events/PM_INST_IMC_MATCH_DISP
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_RMEM_CYC
- /sys/devices/cpu/events/PM_VSU0_SIMPLE_ISSUED
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_RL2L3_SHR
- /sys/devices/cpu/events/PM_VSU_FMA_DOUBLE
- /sys/devices/cpu/events/PM_VSU_4FLOP
- /sys/devices/cpu/events/PM_VSU1_FIN
- /sys/devices/cpu/events/PM_NEST_PAIR1_AND
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_RL2L3_MOD
- /sys/devices/cpu/events/PM_PTEG_FROM_RMEM
- /sys/devices/cpu/events/PM_LSU_LRQ_S0_VALID
- /sys/devices/cpu/events/PM_LSU0_LDF
- /sys/devices/cpu/events/PM_FLUSH_COMPLETION
- /sys/devices/cpu/events/PM_ST_MISS_L1
- /sys/devices/cpu/events/PM_L2_NODE_PUMP
- /sys/devices/cpu/events/PM_INST_FROM_DL2L3_SHR
- /sys/devices/cpu/events/PM_MRK_STALL_CMPLU_CYC
- /sys/devices/cpu/events/PM_VSU1_DENORM
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_L31_SHR_CYC
- /sys/devices/cpu/events/PM_NEST_PAIR0_ADD
- /sys/devices/cpu/events/PM_INST_FROM_L3MISS
- /sys/devices/cpu/events/PM_EE_OFF_EXT_INT
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_DMEM
- /sys/devices/cpu/events/PM_INST_FROM_DL2L3_MOD
- /sys/devices/cpu/events/PM_PMC6_OVERFLOW
- /sys/devices/cpu/events/PM_VSU_2FLOP_DOUBLE
- /sys/devices/cpu/events/PM_TLB_MISS
- /sys/devices/cpu/events/PM_FXU_BUSY
- /sys/devices/cpu/events/PM_L2_RCLD_DISP_FAIL_OTHER
- /sys/devices/cpu/events/PM_LSU_REJECT_LMQ_FULL
- /sys/devices/cpu/events/PM_IC_RELOAD_SHR
- /sys/devices/cpu/events/PM_GRP_MRK
- /sys/devices/cpu/events/PM_MRK_ST_NEST
- /sys/devices/cpu/events/PM_VSU1_FSQRT_FDIV
- /sys/devices/cpu/events/PM_LSU0_FLUSH_LRQ
- /sys/devices/cpu/events/PM_LARX_LSU0
- /sys/devices/cpu/events/PM_IBUF_FULL_CYC
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_DL2L3_SHR_CYC
- /sys/devices/cpu/events/PM_LSU_DC_PREF_STREAM_ALLOC
- /sys/devices/cpu/events/PM_GRP_MRK_CYC
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_RL2L3_SHR_CYC
- /sys/devices/cpu/events/PM_L2_GLOB_GUESS_CORRECT
- /sys/devices/cpu/events/PM_LSU_REJECT_LHS
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_LMEM
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_L3
- /sys/devices/cpu/events/PM_FREQ_DOWN
- /sys/devices/cpu/events/PM_PB_RETRY_NODE_PUMP
- /sys/devices/cpu/events/PM_INST_FROM_RL2L3_SHR
- /sys/devices/cpu/events/PM_MRK_INST_ISSUED
- /sys/devices/cpu/events/PM_PTEG_FROM_L3MISS
- /sys/devices/cpu/events/PM_RUN_PURR
- /sys/devices/cpu/events/PM_MRK_GRP_IC_MISS
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_L3
- /sys/devices/cpu/events/PM_PTEG_FROM_RL2L3_SHR
- /sys/devices/cpu/events/PM_LSU_FLUSH_LRQ
- /sys/devices/cpu/events/PM_MRK_DERAT_MISS_64K
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_DL2L3_MOD
- /sys/devices/cpu/events/PM_L2_ST_MISS
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_L21_SHR
- /sys/devices/cpu/events/PM_LWSYNC
- /sys/devices/cpu/events/PM_LSU0_DC_PREF_STREAM_CONFIRM_STRIDE
- /sys/devices/cpu/events/PM_MRK_LSU_FLUSH_LRQ
- /sys/devices/cpu/events/PM_INST_IMC_MATCH_CMPL
- /sys/devices/cpu/events/PM_NEST_PAIR3_AND
- /sys/devices/cpu/events/PM_PB_RETRY_SYS_PUMP
- /sys/devices/cpu/events/PM_MRK_INST_FIN
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_DL2L3_SHR
- /sys/devices/cpu/events/PM_INST_FROM_L31_MOD
- /sys/devices/cpu/events/PM_MRK_DTLB_MISS_64K
- /sys/devices/cpu/events/PM_LSU_FIN
- /sys/devices/cpu/events/PM_MRK_LSU_REJECT
- /sys/devices/cpu/events/PM_L2_CO_FAIL_BUSY
- /sys/devices/cpu/events/PM_MEM0_WQ_DISP
- /sys/devices/cpu/events/PM_DATA_FROM_L31_MOD
- /sys/devices/cpu/events/PM_THERMAL_WARN
- /sys/devices/cpu/events/PM_VSU0_4FLOP
- /sys/devices/cpu/events/PM_BR_MPRED_CCACHE
- /sys/devices/cpu/events/PM_L1_DEMAND_WRITE
- /sys/devices/cpu/events/PM_FLUSH_BR_MPRED
- /sys/devices/cpu/events/PM_MRK_DTLB_MISS_16G
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_DMEM
- /sys/devices/cpu/events/PM_L2_RCST_DISP
- /sys/devices/cpu/events/PM_LSU_PARTIAL_CDF
- /sys/devices/cpu/events/PM_DISP_CLB_HELD_SB
- /sys/devices/cpu/events/PM_VSU0_FMA_DOUBLE
- /sys/devices/cpu/events/PM_FXU0_BUSY_FXU1_IDLE
- /sys/devices/cpu/events/PM_IC_DEMAND_CYC
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_L21_SHR
- /sys/devices/cpu/events/PM_MRK_LSU_FLUSH_UST
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_L3MISS
- /sys/devices/cpu/events/PM_VSU_DENORM
- /sys/devices/cpu/events/PM_MRK_LSU_PARTIAL_CDF
- /sys/devices/cpu/events/PM_INST_FROM_L21_SHR
- /sys/devices/cpu/events/PM_IC_PREF_WRITE
- /sys/devices/cpu/events/PM_BR_PRED
- /sys/devices/cpu/events/PM_INST_FROM_DMEM
- /sys/devices/cpu/events/PM_IC_PREF_CANCEL_ALL
- /sys/devices/cpu/events/PM_LSU_DC_PREF_STREAM_CONFIRM
- /sys/devices/cpu/events/PM_MRK_LSU_FLUSH_SRQ
- /sys/devices/cpu/events/PM_MRK_FIN_STALL_CYC
- /sys/devices/cpu/events/PM_L2_RCST_DISP_FAIL_OTHER
- /sys/devices/cpu/events/PM_VSU1_DD_ISSUED
- /sys/devices/cpu/events/PM_PTEG_FROM_L31_SHR
- /sys/devices/cpu/events/PM_DATA_FROM_L21_SHR
- /sys/devices/cpu/events/PM_LSU0_NCLD
- /sys/devices/cpu/events/PM_VSU1_4FLOP
- /sys/devices/cpu/events/PM_VSU1_8FLOP
- /sys/devices/cpu/events/PM_VSU_8FLOP
- /sys/devices/cpu/events/PM_LSU_LMQ_SRQ_EMPTY_CYC
- /sys/devices/cpu/events/PM_DTLB_MISS_64K
- /sys/devices/cpu/events/PM_THRD_CONC_RUN_INST
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_L2
- /sys/devices/cpu/events/PM_PB_SYS_PUMP
- /sys/devices/cpu/events/PM_VSU_FIN
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_L31_MOD
- /sys/devices/cpu/events/PM_THRD_PRIO_0_1_CYC
- /sys/devices/cpu/events/PM_DERAT_MISS_64K
- /sys/devices/cpu/events/PM_PMC2_REWIND
- /sys/devices/cpu/events/PM_INST_FROM_L2
- /sys/devices/cpu/events/PM_GRP_BR_MPRED_NONSPEC
- /sys/devices/cpu/events/PM_INST_DISP
- /sys/devices/cpu/events/PM_MEM0_RD_CANCEL_TOTAL
- /sys/devices/cpu/events/PM_LSU0_DC_PREF_STREAM_CONFIRM
- /sys/devices/cpu/events/PM_L1_DCACHE_RELOAD_VALID
- /sys/devices/cpu/events/PM_VSU_SCALAR_DOUBLE_ISSUED
- /sys/devices/cpu/events/PM_L3_PREF_HIT
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_L31_MOD
- /sys/devices/cpu/events/PM_MRK_FXU_FIN
- /sys/devices/cpu/events/PM_PMC4_OVERFLOW
- /sys/devices/cpu/events/PM_MRK_PTEG_FROM_L3
- /sys/devices/cpu/events/PM_LSU0_LMQ_LHR_MERGE
- /sys/devices/cpu/events/PM_BTAC_HIT
- /sys/devices/cpu/events/PM_L3_RD_BUSY
- /sys/devices/cpu/events/PM_LSU0_L1_SW_PREF
- /sys/devices/cpu/events/PM_INST_FROM_L2MISS
- /sys/devices/cpu/events/PM_LSU0_DC_PREF_STREAM_ALLOC
- /sys/devices/cpu/events/PM_L2_ST
- /sys/devices/cpu/events/PM_VSU0_DENORM
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_DL2L3_SHR
- /sys/devices/cpu/events/PM_BR_PRED_CR_TA
- /sys/devices/cpu/events/PM_VSU0_FCONV
- /sys/devices/cpu/events/PM_MRK_LSU_FLUSH_ULD
- /sys/devices/cpu/events/PM_BTAC_MISS
- /sys/devices/cpu/events/PM_MRK_LD_MISS_EXPOSED_CYC_COUNT
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_L2
- /sys/devices/cpu/events/PM_LSU_DCACHE_RELOAD_VALID
- /sys/devices/cpu/events/PM_VSU_FMA
- /sys/devices/cpu/events/PM_LSU0_FLUSH_SRQ
- /sys/devices/cpu/events/PM_LSU1_L1_PREF
- /sys/devices/cpu/events/PM_IOPS_CMPL
- /sys/devices/cpu/events/PM_L2_SYS_PUMP
- /sys/devices/cpu/events/PM_L2_RCLD_BUSY_RC_FULL
- /sys/devices/cpu/events/PM_LSU_LMQ_S0_ALLOC
- /sys/devices/cpu/events/PM_FLUSH_DISP_SYNC
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_DL2L3_MOD_CYC
- /sys/devices/cpu/events/PM_L2_IC_INV
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_L21_MOD_CYC
- /sys/devices/cpu/events/PM_L3_PREF_LDST
- /sys/devices/cpu/events/PM_LSU_SRQ_EMPTY_CYC
- /sys/devices/cpu/events/PM_LSU_LMQ_S0_VALID
- /sys/devices/cpu/events/PM_FLUSH_PARTIAL
- /sys/devices/cpu/events/PM_VSU1_FMA_DOUBLE
- /sys/devices/cpu/events/PM_1PLUS_PPC_DISP
- /sys/devices/cpu/events/PM_DATA_FROM_L2MISS
- /sys/devices/cpu/events/PM_SUSPENDED
- /sys/devices/cpu/events/PM_VSU0_FMA
- /sys/devices/cpu/events/PM_STCX_FAIL
- /sys/devices/cpu/events/PM_VSU0_FSQRT_FDIV_DOUBLE
- /sys/devices/cpu/events/PM_DC_PREF_DST
- /sys/devices/cpu/events/PM_VSU1_SCAL_SINGLE_ISSUED
- /sys/devices/cpu/events/PM_L3_HIT
- /sys/devices/cpu/events/PM_L2_GLOB_GUESS_WRONG
- /sys/devices/cpu/events/PM_MRK_DFU_FIN
- /sys/devices/cpu/events/PM_INST_FROM_L1
- /sys/devices/cpu/events/PM_IC_DEMAND_REQ
- /sys/devices/cpu/events/PM_VSU1_FSQRT_FDIV_DOUBLE
- /sys/devices/cpu/events/PM_VSU1_FMA
- /sys/devices/cpu/events/PM_MRK_LD_MISS_L1
- /sys/devices/cpu/events/PM_VSU0_2FLOP_DOUBLE
- /sys/devices/cpu/events/PM_LSU_DC_PREF_STRIDED_STREAM_CONFIRM
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_L31_SHR
- /sys/devices/cpu/events/PM_MRK_LSU_REJECT_ERAT_MISS
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_L2MISS
- /sys/devices/cpu/events/PM_DATA_FROM_RL2L3_SHR
- /sys/devices/cpu/events/PM_INST_FROM_PREF
- /sys/devices/cpu/events/PM_VSU1_SQ
- /sys/devices/cpu/events/PM_L2_LD_DISP
- /sys/devices/cpu/events/PM_L2_DISP_ALL
- /sys/devices/cpu/events/PM_THRD_GRP_CMPL_BOTH_CYC
- /sys/devices/cpu/events/PM_VSU_FSQRT_FDIV_DOUBLE
- /sys/devices/cpu/events/PM_INST_PTEG_FROM_DL2L3_SHR
- /sys/devices/cpu/events/PM_VSU_1FLOP
- /sys/devices/cpu/events/PM_HV_CYC
- /sys/devices/cpu/events/PM_MRK_LSU_FIN
- /sys/devices/cpu/events/PM_MRK_DATA_FROM_RL2L3_SHR
- /sys/devices/cpu/events/PM_DTLB_MISS_16M
- /sys/devices/cpu/events/PM_LSU1_LMQ_LHR_MERGE
- /sys/devices/cpu/events/PM_IFU_FIN
- /sys/devices/cpu/events/PM_1THRD_CON_RUN_INSTR
- /sys/devices/cpu/events/PM_CMPLU_STALL_COUNT
- /sys/devices/cpu/events/PM_MEM0_PB_RD_CL
- /sys/devices/cpu/events/PM_THRD_1_RUN_CYC
- /sys/devices/cpu/events/PM_THRD_2_CONC_RUN_INSTR
- /sys/devices/cpu/events/PM_THRD_2_RUN_CYC
- /sys/devices/cpu/events/PM_THRD_3_CONC_RUN_INST
- /sys/devices/cpu/events/PM_THRD_3_RUN_CYC
- /sys/devices/cpu/events/PM_THRD_4_CONC_RUN_INST
- /sys/devices/cpu/events/PM_THRD_4_RUN_CYC
-
-Date: 2013/01/08
-
-Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
- Linux Powerpc mailing list <linuxppc-dev@ozlabs.org>
-
-Description: POWER-systems specific performance monitoring events
-
- A collection of performance monitoring events that may be
- supported by the POWER CPU. These events can be monitored
- using the 'perf(1)' tool.
-
- These events may not be supported by other CPUs.
-
- The contents of each file would look like:
-
- event=0xNNNN
-
- where 'N' is a hex digit and the number '0xNNNN' shows the
- "raw code" for the perf event identified by the file's
- "basename".
-
- Further, multiple terms like 'event=0xNNNN' can be specified
- and separated with comma. All available terms are defined in
- the /sys/bus/event_source/devices/<dev>/format file.
-
What: /sys/bus/event_source/devices/<pmu>/events/<event>
Date: 2014/02/24
Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
--
1.7.9.5
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox