From: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
To: Ashutosh Dixit <ashutosh.dixit@intel.com>
Cc: Harish Chegondi <harish.chegondi@intel.com>,
intel-xe@lists.freedesktop.org, gzadicario@habana.ai,
ogabbay@habana.ai
Subject: Re: [Intel-xe] [PATCH 1/1] drm/xe/uapi: "Perf" layer to support multiple perf counter stream types
Date: Mon, 30 Oct 2023 10:09:29 -0700 [thread overview]
Message-ID: <ZT/jSWWLMP0U8jB5@unerlige-ril> (raw)
In-Reply-To: <20231028033910.2479157-2-ashutosh.dixit@intel.com>
On Fri, Oct 27, 2023 at 08:39:10PM -0700, Ashutosh Dixit wrote:
>In Xe, the plan is to support multiple types of perf counter streams (OA is
>only one type of these streams). Rather than introduce NxM ioctls for
>these (N perf streams with M ioctl's per perf stream), we decide to
>multiplex these (N different stream types and the M ops for each of these
>stream types) through a single PERF ioctl. This multiplexing is the purpose
>of the PERF layer.
>
>In addition to PERF DRM ioctl's, another set of ioctl's on the PERF fd are
>defined. These are expected to be common to different PERF stream types and
>therefore defined at the PERF layer itself.
>
>v2: Add param_size to 'struct drm_xe_perf_param' (Umesh)
>v3: Rename 'enum drm_xe_perf_ops' to
> 'enum drm_xe_perf_ioctls' (Guy Zadicario)
> Add DRM_ prefix to ioctl names to indicate uapi names
>v4: Add 'enum drm_xe_perf_op' previously missed out (Guy Zadicario)
>v5: Squash the ops and PERF layer patches into a single patch (Umesh)
> Remove param_size from struct 'drm_xe_perf_param' (Umesh)
>
>Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
lgtm from OA perspective
Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Thanks,
Umesh
>---
> drivers/gpu/drm/xe/Makefile | 1 +
> drivers/gpu/drm/xe/xe_device.c | 2 ++
> drivers/gpu/drm/xe/xe_perf.c | 21 ++++++++++++
> drivers/gpu/drm/xe/xe_perf.h | 16 +++++++++
> include/uapi/drm/xe_drm.h | 59 ++++++++++++++++++++++++++++++++++
> 5 files changed, 99 insertions(+)
> create mode 100644 drivers/gpu/drm/xe/xe_perf.c
> create mode 100644 drivers/gpu/drm/xe/xe_perf.h
>
>diff --git a/drivers/gpu/drm/xe/Makefile b/drivers/gpu/drm/xe/Makefile
>index cee57681732d5..87c57f6c927e9 100644
>--- a/drivers/gpu/drm/xe/Makefile
>+++ b/drivers/gpu/drm/xe/Makefile
>@@ -89,6 +89,7 @@ xe-y += xe_bb.o \
> xe_pat.o \
> xe_pci.o \
> xe_pcode.o \
>+ xe_perf.o \
> xe_pm.o \
> xe_preempt_fence.o \
> xe_pt.o \
>diff --git a/drivers/gpu/drm/xe/xe_device.c b/drivers/gpu/drm/xe/xe_device.c
>index 8341acf66e5f9..4b5a0c504eda1 100644
>--- a/drivers/gpu/drm/xe/xe_device.c
>+++ b/drivers/gpu/drm/xe/xe_device.c
>@@ -28,6 +28,7 @@
> #include "xe_module.h"
> #include "xe_pat.h"
> #include "xe_pcode.h"
>+#include "xe_perf.h"
> #include "xe_pm.h"
> #include "xe_query.h"
> #include "xe_tile.h"
>@@ -127,6 +128,7 @@ static const struct drm_ioctl_desc xe_ioctls[] = {
> DRM_IOCTL_DEF_DRV(XE_WAIT_USER_FENCE, xe_wait_user_fence_ioctl,
> DRM_RENDER_ALLOW),
> DRM_IOCTL_DEF_DRV(XE_VM_MADVISE, xe_vm_madvise_ioctl, DRM_RENDER_ALLOW),
>+ DRM_IOCTL_DEF_DRV(XE_PERF, xe_perf_ioctl, DRM_RENDER_ALLOW),
> };
>
> static const struct file_operations xe_driver_fops = {
>diff --git a/drivers/gpu/drm/xe/xe_perf.c b/drivers/gpu/drm/xe/xe_perf.c
>new file mode 100644
>index 0000000000000..a130076b59aa2
>--- /dev/null
>+++ b/drivers/gpu/drm/xe/xe_perf.c
>@@ -0,0 +1,21 @@
>+// SPDX-License-Identifier: MIT
>+/*
>+ * Copyright © 2023 Intel Corporation
>+ */
>+
>+#include <linux/errno.h>
>+
>+#include "xe_perf.h"
>+
>+int xe_perf_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
>+{
>+ struct drm_xe_perf_param *arg = data;
>+
>+ if (arg->extensions)
>+ return -EINVAL;
>+
>+ switch (arg->perf_type) {
>+ default:
>+ return -EINVAL;
>+ }
>+}
>diff --git a/drivers/gpu/drm/xe/xe_perf.h b/drivers/gpu/drm/xe/xe_perf.h
>new file mode 100644
>index 0000000000000..254cc7cf49fef
>--- /dev/null
>+++ b/drivers/gpu/drm/xe/xe_perf.h
>@@ -0,0 +1,16 @@
>+/* SPDX-License-Identifier: MIT */
>+/*
>+ * Copyright © 2023 Intel Corporation
>+ */
>+
>+#ifndef _XE_PERF_H_
>+#define _XE_PERF_H_
>+
>+#include <drm/xe_drm.h>
>+
>+struct drm_device;
>+struct drm_file;
>+
>+int xe_perf_ioctl(struct drm_device *dev, void *data, struct drm_file *file);
>+
>+#endif
>diff --git a/include/uapi/drm/xe_drm.h b/include/uapi/drm/xe_drm.h
>index 24bf8f0f52e82..3eb4b20159824 100644
>--- a/include/uapi/drm/xe_drm.h
>+++ b/include/uapi/drm/xe_drm.h
>@@ -110,6 +110,7 @@ struct xe_user_extension {
> #define DRM_XE_WAIT_USER_FENCE 0x0a
> #define DRM_XE_VM_MADVISE 0x0b
> #define DRM_XE_EXEC_QUEUE_GET_PROPERTY 0x0c
>+#define DRM_XE_PERF 0x0d
>
> /* Must be kept compact -- no holes */
> #define DRM_IOCTL_XE_DEVICE_QUERY DRM_IOWR(DRM_COMMAND_BASE + DRM_XE_DEVICE_QUERY, struct drm_xe_device_query)
>@@ -125,6 +126,7 @@ struct xe_user_extension {
> #define DRM_IOCTL_XE_EXEC_QUEUE_SET_PROPERTY DRM_IOW(DRM_COMMAND_BASE + DRM_XE_EXEC_QUEUE_SET_PROPERTY, struct drm_xe_exec_queue_set_property)
> #define DRM_IOCTL_XE_WAIT_USER_FENCE DRM_IOWR(DRM_COMMAND_BASE + DRM_XE_WAIT_USER_FENCE, struct drm_xe_wait_user_fence)
> #define DRM_IOCTL_XE_VM_MADVISE DRM_IOW(DRM_COMMAND_BASE + DRM_XE_VM_MADVISE, struct drm_xe_vm_madvise)
>+#define DRM_IOCTL_XE_PERF DRM_IOW(DRM_COMMAND_BASE + DRM_XE_PERF, struct drm_xe_perf_param)
>
> /** struct drm_xe_engine_class_instance - instance of an engine class */
> struct drm_xe_engine_class_instance {
>@@ -1074,6 +1076,63 @@ struct drm_xe_vm_madvise {
> #define XE_PMU_MEDIA_GROUP_BUSY(gt) ___XE_PMU_OTHER(gt, 3)
> #define XE_PMU_ANY_ENGINE_GROUP_BUSY(gt) ___XE_PMU_OTHER(gt, 4)
>
>+/**
>+ * enum drm_xe_perf_type - Perf stream types
>+ */
>+enum drm_xe_perf_type {
>+ DRM_XE_PERF_TYPE_MAX,
>+};
>+
>+/**
>+ * enum drm_xe_perf_op - Perf stream ops
>+ */
>+enum drm_xe_perf_op {
>+ /** @DRM_XE_PERF_OP_STREAM_OPEN: Open a perf counter stream */
>+ DRM_XE_PERF_OP_STREAM_OPEN,
>+
>+ /** @DRM_XE_PERF_OP_ADD_CONFIG: Add perf stream config */
>+ DRM_XE_PERF_OP_ADD_CONFIG,
>+
>+ /** @DRM_XE_PERF_OP_REMOVE_CONFIG: Remove perf stream config */
>+ DRM_XE_PERF_OP_REMOVE_CONFIG,
>+};
>+
>+/**
>+ * struct drm_xe_perf_param - Perf layer param
>+ *
>+ * The perf layer enables multiplexing perf counter streams of multiple
>+ * types. The actual params for a particular stream operation are supplied
>+ * via the @param pointer (use __copy_from_user to get these params).
>+ */
>+struct drm_xe_perf_param {
>+ /** @extensions: Pointer to the first extension struct, if any */
>+ __u64 extensions;
>+ /** @perf_type: Perf stream type, of enum @drm_xe_perf_type */
>+ __u64 perf_type;
>+ /** @perf_op: Perf op, of enum @drm_xe_perf_op */
>+ __u64 perf_op;
>+ /** @param: Pointer to actual stream params */
>+ __u64 param;
>+};
>+
>+/**
>+ * enum drm_xe_perf_ioctls - Perf fd ioctl's
>+ */
>+enum drm_xe_perf_ioctls {
>+ /**
>+ * @DRM_XE_PERF_IOCTL_ENABLE: Enable data capture for a stream that
>+ * was e.g. either opened in a disabled state or was disabled via
>+ * DRM_XE_PERF_IOCTL_DISABLE
>+ */
>+ DRM_XE_PERF_IOCTL_ENABLE = _IO('i', 0x0),
>+
>+ /** @DRM_XE_PERF_IOCTL_DISABLE: Disable data capture for a stream */
>+ DRM_XE_PERF_IOCTL_DISABLE = _IO('i', 0x1),
>+
>+ /** @DRM_XE_PERF_IOCTL_CONFIG: Change stream configuration */
>+ DRM_XE_PERF_IOCTL_CONFIG = _IO('i', 0x2),
>+};
>+
> #if defined(__cplusplus)
> }
> #endif
>--
>2.41.0
>
next prev parent reply other threads:[~2023-10-30 17:09 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-10-28 3:39 [Intel-xe] [PATCH v3 0/1] Xe PERF layer Ashutosh Dixit
2023-10-28 3:39 ` [Intel-xe] [PATCH 1/1] drm/xe/uapi: "Perf" layer to support multiple perf counter stream types Ashutosh Dixit
2023-10-30 17:09 ` Umesh Nerlige Ramappa [this message]
2023-10-28 3:41 ` [Intel-xe] ✓ CI.Patch_applied: success for Xe PERF layer Patchwork
2023-10-28 3:41 ` [Intel-xe] ✗ CI.checkpatch: warning " Patchwork
2023-10-28 3:42 ` [Intel-xe] ✓ CI.KUnit: success " Patchwork
2023-10-28 3:49 ` [Intel-xe] ✓ CI.Build: " Patchwork
2023-10-28 3:50 ` [Intel-xe] ✗ CI.Hooks: failure " Patchwork
2023-10-28 3:51 ` [Intel-xe] ✓ CI.checksparse: success " Patchwork
2023-10-28 4:28 ` [Intel-xe] ✓ CI.BAT: " Patchwork
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ZT/jSWWLMP0U8jB5@unerlige-ril \
--to=umesh.nerlige.ramappa@intel.com \
--cc=ashutosh.dixit@intel.com \
--cc=gzadicario@habana.ai \
--cc=harish.chegondi@intel.com \
--cc=intel-xe@lists.freedesktop.org \
--cc=ogabbay@habana.ai \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox