Re: [PATCH 02/10] drm/ttm: flip over the range manager to self allocated nodes

Thomas Hellström (Intel) <thomas_os@xxxxxxxxxxxx> · Wed, 2 Jun 2021 14:33:11 +0200

On 6/2/21 2:11 PM, Christian König wrote:
Am 02.06.21 um 13:44 schrieb Thomas Hellström (Intel):

On 6/2/21 12:09 PM, Christian König wrote:
Start with the range manager to make the resource object the base
class for the allocated nodes.

While at it cleanup a lot of the code around that.

Signed-off-by: Christian König <christian.koenig@xxxxxxx>
Reviewed-by: Matthew Auld <matthew.auld@xxxxxxxxx>
---
  drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c |  1 +
  drivers/gpu/drm/drm_gem_vram_helper.c   |  2 +
  drivers/gpu/drm/nouveau/nouveau_ttm.c   |  2 +
  drivers/gpu/drm/qxl/qxl_ttm.c           |  1 +
  drivers/gpu/drm/radeon/radeon_ttm.c     |  1 +
  drivers/gpu/drm/ttm/ttm_range_manager.c | 56 
++++++++++++++++++-------
  drivers/gpu/drm/ttm/ttm_resource.c      | 26 ++++++++----
  include/drm/ttm/ttm_bo_driver.h         | 26 ------------
  include/drm/ttm/ttm_range_manager.h     | 43 +++++++++++++++++++
  include/drm/ttm/ttm_resource.h          |  3 ++
  10 files changed, 111 insertions(+), 50 deletions(-)
  create mode 100644 include/drm/ttm/ttm_range_manager.h

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c 
b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 69db89261650..df1f185faae9 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -45,6 +45,7 @@
  #include <drm/ttm/ttm_bo_api.h>
  #include <drm/ttm/ttm_bo_driver.h>
  #include <drm/ttm/ttm_placement.h>
+#include <drm/ttm/ttm_range_manager.h>
    #include <drm/amdgpu_drm.h>
  diff --git a/drivers/gpu/drm/drm_gem_vram_helper.c 
b/drivers/gpu/drm/drm_gem_vram_helper.c
index 83e7258c7f90..17a4c5d47b6a 100644
--- a/drivers/gpu/drm/drm_gem_vram_helper.c
+++ b/drivers/gpu/drm/drm_gem_vram_helper.c
@@ -17,6 +17,8 @@
  #include <drm/drm_prime.h>
  #include <drm/drm_simple_kms_helper.h>
  +#include <drm/ttm/ttm_range_manager.h>
+
  static const struct drm_gem_object_funcs drm_gem_vram_object_funcs;
    /**
diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c 
b/drivers/gpu/drm/nouveau/nouveau_ttm.c
index 65430912ff72..b08b8efeefba 100644
--- a/drivers/gpu/drm/nouveau/nouveau_ttm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c
@@ -26,6 +26,8 @@
  #include <linux/limits.h>
  #include <linux/swiotlb.h>
  +#include <drm/ttm/ttm_range_manager.h>
+
  #include "nouveau_drv.h"
  #include "nouveau_gem.h"
  #include "nouveau_mem.h"
diff --git a/drivers/gpu/drm/qxl/qxl_ttm.c 
b/drivers/gpu/drm/qxl/qxl_ttm.c
index 8aa87b8edb9c..19fd39d9a00c 100644
--- a/drivers/gpu/drm/qxl/qxl_ttm.c
+++ b/drivers/gpu/drm/qxl/qxl_ttm.c
@@ -32,6 +32,7 @@
  #include <drm/ttm/ttm_bo_api.h>
  #include <drm/ttm/ttm_bo_driver.h>
  #include <drm/ttm/ttm_placement.h>
+#include <drm/ttm/ttm_range_manager.h>
    #include "qxl_drv.h"
  #include "qxl_object.h"
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c 
b/drivers/gpu/drm/radeon/radeon_ttm.c
index cdffa9b65108..ad2a5a791bba 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -45,6 +45,7 @@
  #include <drm/ttm/ttm_bo_api.h>
  #include <drm/ttm/ttm_bo_driver.h>
  #include <drm/ttm/ttm_placement.h>
+#include <drm/ttm/ttm_range_manager.h>
    #include "radeon_reg.h"
  #include "radeon.h"
diff --git a/drivers/gpu/drm/ttm/ttm_range_manager.c 
b/drivers/gpu/drm/ttm/ttm_range_manager.c
index b9d5da6e6a81..ce5d07ca384c 100644
--- a/drivers/gpu/drm/ttm/ttm_range_manager.c
+++ b/drivers/gpu/drm/ttm/ttm_range_manager.c
@@ -29,12 +29,13 @@
   * Authors: Thomas Hellstrom <thellstrom-at-vmware-dot-com>
   */
  -#include <drm/ttm/ttm_bo_driver.h>
+#include <drm/ttm/ttm_device.h>
  #include <drm/ttm/ttm_placement.h>
+#include <drm/ttm/ttm_range_manager.h>
+#include <drm/ttm/ttm_bo_api.h>
  #include <drm/drm_mm.h>
  #include <linux/slab.h>
  #include <linux/spinlock.h>
-#include <linux/module.h>
    /*
   * Currently we use a spinlock for the lock, but a mutex *may* be
@@ -60,8 +61,8 @@ static int ttm_range_man_alloc(struct 
ttm_resource_manager *man,
                     struct ttm_resource *mem)
  {
      struct ttm_range_manager *rman = to_range_manager(man);
+    struct ttm_range_mgr_node *node;
      struct drm_mm *mm = &rman->mm;
-    struct drm_mm_node *node;
      enum drm_mm_insert_mode mode;
      unsigned long lpfn;
      int ret;
@@ -70,7 +71,7 @@ static int ttm_range_man_alloc(struct 
ttm_resource_manager *man,
      if (!lpfn)
          lpfn = man->size;
  -    node = kzalloc(sizeof(*node), GFP_KERNEL);
+    node = kzalloc(struct_size(node, mm_nodes, 1), GFP_KERNEL);

I'm still a bit confused  about the situation where a driver wants to 
attach private data to a struct ttm_resource without having to 
re-implement its own range manager?

Could be cached sg-tables, list of GPU bindings etc. Wouldn't work 
with the above unless we have a void *driver_private member on the 
struct ttm_resource. Is that the plan going forward here? Or that the 
driver actually does the re-implementation?

I don't really understand your concern here. The basic idea is that 
drivers use ttm_resource as a base class for their own implementation.

See for example how nouveau does that:

struct nouveau_mem {
        struct ttm_resource base;
        struct nouveau_cli *cli;
        u8 kind;
        u8 comp;
        struct nvif_mem mem;
        struct nvif_vma vma[2];
};

The range manager is helping driver specific resource managers which 
want to implement something drm_mm_nodes based. E.g. amdgpu_gtt_mgr 
and amdgpu_vram_mgr, but it can also be used stand alone.

The ttm_range_mgr_node can then be used as base class for this 
functionality. I already want to move some more code from 
amdgpu_vram_mgr.c into the range manager, but that is just minor 
cleanup work.

Sure but if you embed a ttm_range_mgr_node in your struct i915_resource, 
and wanted to use the ttm range manager for it, it would allocate a 
struct ttm_range_mgr_node rather than a struct i915_resource? Or am I 
missing something?

/Thomas



Regards,
Christian.


Thanks,

Thomas








