All of lore.kernel.org
 help / color / mirror / Atom feed
From: Chris Wilson <chris@chris-wilson.co.uk>
To: intel-gfx@lists.freedesktop.org
Subject: [PATCH 32/40] drm/i915: Introduce the i915_user_extension_method
Date: Wed, 19 Sep 2018 20:55:36 +0100	[thread overview]
Message-ID: <20180919195544.1511-32-chris@chris-wilson.co.uk> (raw)
In-Reply-To: <20180919195544.1511-1-chris@chris-wilson.co.uk>

An idea for extending uABI inspired by Vulkan's extension chains.
Instead of expanding the data struct for each ioctl every time we need
to add a new feature, define an extension chain instead. As we add
optional interfaces to control the ioctl, we define a new extension
struct that can be linked into the ioctl data only when required by the
user. The key advantage being able to ignore large control structs for
optional interfaces/extensions, while being able to process them in a
consistent manner.

In comparison to other extensible ioctls, the key difference is the
use of a linked chain of extension structs vs an array of tagged
pointers. For example,

struct drm_amdgpu_cs_chunk {
        __u32           chunk_id;
        __u32           length_dw;
        __u64           chunk_data;
};

struct drm_amdgpu_cs_in {
        __u32           ctx_id;
        __u32           bo_list_handle;
        __u32           num_chunks;
        __u32           _pad;
        __u64           chunks;
};

allows userspace to pass in array of pointers to extension structs, but
must therefore keep constructing that array along side the command stream.
In dynamic situations like that, a linked list is preferred and does not
similar from extra cache line misses as the extension structs themselves
must still be loaded separate to the chunks array.

v2: Apply the tail call optimisation directly to nip the worry of stack
overflow in the bud.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/Makefile               |  1 +
 drivers/gpu/drm/i915/i915_user_extensions.c | 37 +++++++++++++++++++++
 drivers/gpu/drm/i915/i915_user_extensions.h | 20 +++++++++++
 include/uapi/drm/i915_drm.h                 | 20 +++++++++++
 4 files changed, 78 insertions(+)
 create mode 100644 drivers/gpu/drm/i915/i915_user_extensions.c
 create mode 100644 drivers/gpu/drm/i915/i915_user_extensions.h

diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index 4654ef481b45..6ffeb60392f5 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -45,6 +45,7 @@ i915-y := i915_drv.o \
 	  i915_syncmap.o \
 	  i915_sw_fence.o \
 	  i915_sysfs.o \
+	  i915_user_extensions.o \
 	  intel_csr.o \
 	  intel_device_info.o \
 	  intel_pm.o \
diff --git a/drivers/gpu/drm/i915/i915_user_extensions.c b/drivers/gpu/drm/i915/i915_user_extensions.c
new file mode 100644
index 000000000000..00622f8ac468
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_user_extensions.c
@@ -0,0 +1,37 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2018 Intel Corporation
+ */
+
+#include <linux/uaccess.h>
+#include <uapi/drm/i915_drm.h>
+
+#include "i915_user_extensions.h"
+
+int i915_user_extensions(struct i915_user_extension __user *ext,
+			 const i915_user_extension_fn *tbl,
+			 unsigned long count,
+			 void *data)
+{
+	while (ext) {
+		int err;
+		u64 x;
+
+		if (get_user(x, &ext->name))
+			return -EFAULT;
+
+		err = -EINVAL;
+		if (x < count && tbl[x])
+			err = tbl[x](ext, data);
+		if (err)
+			return err;
+
+		if (get_user(x, &ext->next_extension))
+			return -EFAULT;
+
+		ext = u64_to_user_ptr(x);
+	}
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/i915/i915_user_extensions.h b/drivers/gpu/drm/i915/i915_user_extensions.h
new file mode 100644
index 000000000000..313a510b068a
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_user_extensions.h
@@ -0,0 +1,20 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2018 Intel Corporation
+ */
+
+#ifndef I915_USER_EXTENSIONS_H
+#define I915_USER_EXTENSIONS_H
+
+struct i915_user_extension;
+
+typedef int (*i915_user_extension_fn)(struct i915_user_extension __user *ext,
+				      void *data);
+
+int i915_user_extensions(struct i915_user_extension __user *ext,
+			 const i915_user_extension_fn *tbl,
+			 unsigned long count,
+			 void *data);
+
+#endif /* I915_USER_EXTENSIONS_H */
diff --git a/include/uapi/drm/i915_drm.h b/include/uapi/drm/i915_drm.h
index a4446f452040..4fa7e54501cc 100644
--- a/include/uapi/drm/i915_drm.h
+++ b/include/uapi/drm/i915_drm.h
@@ -62,6 +62,26 @@ extern "C" {
 #define I915_ERROR_UEVENT		"ERROR"
 #define I915_RESET_UEVENT		"RESET"
 
+/*
+ * i915_user_extension: Base class for defining a chain of extensions
+ *
+ * Many interfaces need to grow over time. In most cases we can simply
+ * extend the struct and have userspace pass in more data. Another option,
+ * as demonstrated by Vulkan's approach to providing extensions for forward
+ * and backward compatibilty, is to use a list of optional structs to
+ * provide those extra details.
+ *
+ * The key advantage to using an extension chain is that it allows us to
+ * redefine the interface more easily than an ever growing struct of
+ * increasing complexity, and for large parts of that interface to be
+ * entirely optional. The downside is more pointer chasing; chasing across
+ * the __user boundary with pointers encapsulated inside u64.
+ */
+struct i915_user_extension {
+	__u64 next_extension;
+	__u64 name;
+};
+
 /*
  * MOCS indexes used for GPU surfaces, defining the cacheability of the
  * surface data and the coherency for this data wrt. CPU vs. GPU accesses.
-- 
2.19.0

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

  parent reply	other threads:[~2018-09-19 19:56 UTC|newest]

Thread overview: 106+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-09-19 19:55 [PATCH 01/40] drm: Use default dma_fence hooks where possible for null syncobj Chris Wilson
2018-09-19 19:55 ` [PATCH 02/40] drm: Fix syncobj handing of schedule() returning 0 Chris Wilson
2018-09-20 14:13   ` Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 03/40] drm/i915/selftests: Live tests emit requests and so require rpm Chris Wilson
2018-09-20 14:01   ` Tvrtko Ursulin
2018-09-20 14:47     ` Chris Wilson
2018-09-19 19:55 ` [PATCH 04/40] drm/i915: Park the GPU on module load Chris Wilson
2018-09-20 14:02   ` Tvrtko Ursulin
2018-09-20 14:52     ` Chris Wilson
2018-09-19 19:55 ` [PATCH 05/40] drm/i915: Handle incomplete Z_FINISH for compressed error states Chris Wilson
2018-09-19 19:55   ` Chris Wilson
2018-09-19 19:55 ` [PATCH 06/40] drm/i915: Clear the error PTE just once on finish Chris Wilson
2018-09-19 19:55 ` [PATCH 07/40] drm/i915: Cache the error string Chris Wilson
2018-09-19 19:55 ` [PATCH 08/40] drm/i915/execlists: Avoid kicking priority on the current context Chris Wilson
2018-09-19 19:55 ` [PATCH 09/40] drm/i915/selftests: Free the batch along the contexts error path Chris Wilson
2018-09-20  8:30   ` Mika Kuoppala
2018-09-20  8:36     ` Chris Wilson
2018-09-20  9:19       ` Mika Kuoppala
2018-09-19 19:55 ` [PATCH 10/40] drm/i915/selftests: Basic stress test for rapid context switching Chris Wilson
2018-09-20 10:38   ` Mika Kuoppala
2018-09-20 10:46     ` Chris Wilson
2018-09-19 19:55 ` [PATCH 11/40] drm/i915/execlists: Onion unwind for logical_ring_init() failure Chris Wilson
2018-09-20 14:18   ` Mika Kuoppala
2018-09-20 14:21   ` Tvrtko Ursulin
2018-09-20 19:59     ` Chris Wilson
2018-09-21 10:00       ` Tvrtko Ursulin
2018-09-21 10:01         ` Chris Wilson
2018-09-19 19:55 ` [PATCH 12/40] drm/i915/execlists: Assert the queue is non-empty on unsubmitting Chris Wilson
2018-09-24  9:07   ` Tvrtko Ursulin
2018-09-25  7:41     ` Chris Wilson
2018-09-25  8:51       ` Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 13/40] drm/i915: Reserve some priority bits for internal use Chris Wilson
2018-09-24  9:12   ` Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 14/40] drm/i915: Combine multiple internal plists into the same i915_priolist bucket Chris Wilson
2018-09-24 10:25   ` Tvrtko Ursulin
2018-09-25  7:55     ` Chris Wilson
2018-09-19 19:55 ` [PATCH 15/40] drm/i915: Priority boost for new clients Chris Wilson
2018-09-24 10:29   ` Tvrtko Ursulin
2018-09-25  8:01     ` Chris Wilson
2018-09-25  8:26       ` Chris Wilson
2018-09-25  8:57         ` Tvrtko Ursulin
2018-09-25  9:06           ` Chris Wilson
2018-09-25  9:08             ` Tvrtko Ursulin
2018-09-25 11:20         ` Michal Wajdeczko
2018-09-19 19:55 ` [PATCH 16/40] drm/i915: Pull scheduling under standalone lock Chris Wilson
2018-09-24 11:19   ` Tvrtko Ursulin
2018-09-25  8:19     ` Chris Wilson
2018-09-25  9:01       ` Tvrtko Ursulin
2018-09-25  9:10         ` Chris Wilson
2018-09-25  9:19           ` Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 17/40] drm/i915: Priority boost for waiting clients Chris Wilson
2018-09-24 11:29   ` Tvrtko Ursulin
2018-09-25  9:00     ` Chris Wilson
2018-09-25  9:07       ` Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 18/40] drm/i915: Report the number of closed vma held by each context in debugfs Chris Wilson
2018-09-24 11:57   ` Tvrtko Ursulin
2018-09-25 12:20     ` Chris Wilson
2018-09-19 19:55 ` [PATCH 19/40] drm/i915: Remove debugfs/i915_ppgtt_info Chris Wilson
2018-09-24 12:03   ` Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 20/40] drm/i915: Track all held rpm wakerefs Chris Wilson
2018-09-19 19:55 ` [PATCH 21/40] drm/i915: Markup paired operations on wakerefs Chris Wilson
2018-09-19 19:55 ` [PATCH 22/40] drm/i915: Syntatic sugar for using intel_runtime_pm Chris Wilson
2018-09-24 12:08   ` Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 23/40] drm/i915: Markup paired operations on display power domains Chris Wilson
2018-09-19 19:55 ` [PATCH 24/40] drm/i915: Track the wakeref used to initialise " Chris Wilson
2018-09-19 19:55 ` [PATCH 25/40] drm/i915/dp: Markup pps lock power well Chris Wilson
2018-09-19 19:55 ` [PATCH 26/40] drm/i915: Complain if hsw_get_pipe_config acquires the same power well twice Chris Wilson
2018-09-19 19:55 ` [PATCH 27/40] drm/i915: Mark up Ironlake ips with rpm wakerefs Chris Wilson
2018-09-19 19:55 ` [PATCH 28/40] drm/i915: Serialise concurrent calls to i915_gem_set_wedged() Chris Wilson
2018-09-19 19:55 ` [PATCH 29/40] drm/i915: Differentiate between ggtt->mutex and ppgtt->mutex Chris Wilson
2018-09-24 13:04   ` Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 30/40] drm/i915: Pull all the reset functionality together into i915_reset.c Chris Wilson
2018-09-19 19:55 ` [PATCH 31/40] drm/i915: Make all GPU resets atomic Chris Wilson
2018-09-19 19:55 ` Chris Wilson [this message]
2018-09-24 13:20   ` [PATCH 32/40] drm/i915: Introduce the i915_user_extension_method Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 33/40] drm/i915: Extend CREATE_CONTEXT to allow inheritance ala clone() Chris Wilson
2018-09-24 17:22   ` Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 34/40] drm/i915: Allow contexts to share a single timeline across all engines Chris Wilson
2018-09-25  8:45   ` Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 35/40] drm/i915: Fix I915_EXEC_RING_MASK Chris Wilson
2018-09-25  8:46   ` [Intel-gfx] " Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 36/40] drm/i915: Re-arrange execbuf so context is known before engine Chris Wilson
2018-09-19 19:55 ` [PATCH 37/40] drm/i915: Allow a context to define its set of engines Chris Wilson
2018-09-27 11:28   ` Tvrtko Ursulin
2018-09-28 20:22     ` Chris Wilson
2018-10-01  8:30       ` Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 38/40] drm/i915/execlists: Flush the CS events before unpinning Chris Wilson
2018-10-01 10:51   ` Tvrtko Ursulin
2018-10-01 11:06     ` Chris Wilson
2018-10-01 13:15       ` Tvrtko Ursulin
2018-10-01 13:26         ` Chris Wilson
2018-10-01 14:03           ` Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 39/40] drm/i915/execlists: Refactor out can_merge_rq() Chris Wilson
2018-09-27 11:32   ` Tvrtko Ursulin
2018-09-28 20:11     ` Chris Wilson
2018-10-01  8:14       ` Tvrtko Ursulin
2018-10-01  8:18         ` Chris Wilson
2018-10-01 10:18           ` Tvrtko Ursulin
2018-09-19 19:55 ` [PATCH 40/40] drm/i915: Load balancing across a virtual engine Chris Wilson
2018-10-01 11:37   ` Tvrtko Ursulin
2018-09-19 21:54 ` ✗ Fi.CI.CHECKPATCH: warning for series starting with [01/40] drm: Use default dma_fence hooks where possible for null syncobj Patchwork
2018-09-19 22:08 ` ✗ Fi.CI.SPARSE: " Patchwork
2018-09-19 22:17 ` ✗ Fi.CI.BAT: failure " Patchwork
2018-09-20 13:34 ` [PATCH 01/40] " Tvrtko Ursulin
2018-09-20 13:40   ` Chris Wilson
2018-09-20 13:54     ` Tvrtko Ursulin

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=20180919195544.1511-32-chris@chris-wilson.co.uk \
    --to=chris@chris-wilson.co.uk \
    --cc=intel-gfx@lists.freedesktop.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.