On Thu, Jul 18, 2024 at 06:12:30PM +1000, Ben Skeggs wrote:
> On 10/7/24 02:38, Danilo Krummrich wrote:
>
> > On Fri, Jul 05, 2024 at 04:37:18AM +1000, Ben Skeggs wrote:
> > > The only remaining nouveau_drm.master struct member that's being used is
> > > the mutex that protects its object tree. Move that into nouveau_drm and
> > > remove nouveau_drm.master entirely.
> > >
> > > Another patch series will make it possible to also remove the mutex, but
> > > it's still required for the moment.
> > It would be good to also mention what we're left with to protect and how we
> > get
> > rid of that later on.
> I already mention what's left to protect? I've modified the commit message
> to mention the remove-ioctl series regardless.
Indeed, that is fine. I think I just read over this, sorry for the noise.
> >
> > > Signed-off-by: Ben Skeggs <bske...@nvidia.com>
> > > ---
> > > drivers/gpu/drm/nouveau/nouveau_drm.c | 39 +++++++++++----------------
> > > drivers/gpu/drm/nouveau/nouveau_drv.h | 3 ++-
> > > drivers/gpu/drm/nouveau/nouveau_mem.c | 12 ++++-----
> > > 3 files changed, 23 insertions(+), 31 deletions(-)
> > >
> > > diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c
> > > b/drivers/gpu/drm/nouveau/nouveau_drm.c
> > > index c6c9a528783a..85214d35fd5d 100644
> > > --- a/drivers/gpu/drm/nouveau/nouveau_drm.c
> > > +++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
> > > @@ -208,11 +208,9 @@ nouveau_cli_fini(struct nouveau_cli *cli)
> > > nvif_mmu_dtor(&cli->mmu);
> > > cli->device.object.map.ptr = NULL;
> > > nvif_device_dtor(&cli->device);
> > > - if (cli != &cli->drm->master) {
> > > - mutex_lock(&cli->drm->master.lock);
> > > - nvif_client_dtor(&cli->base);
> > > - mutex_unlock(&cli->drm->master.lock);
> > > - }
> > > + mutex_lock(&cli->drm->client_mutex);
> > > + nvif_client_dtor(&cli->base);
> > > + mutex_unlock(&cli->drm->client_mutex);
> > > }
> > > static int
> > > @@ -245,11 +243,9 @@ nouveau_cli_init(struct nouveau_drm *drm, const char
> > > *sname,
> > > INIT_LIST_HEAD(&cli->worker);
> > > mutex_init(&cli->lock);
> > > - if (cli != &drm->master) {
> > > - mutex_lock(&drm->master.lock);
> > > - ret = nvif_client_ctor(&drm->master.base, cli->name,
> > > &cli->base);
> > > - mutex_unlock(&drm->master.lock);
> > > - }
> > > + mutex_lock(&drm->client_mutex);
> > > + ret = nvif_client_ctor(&drm->_client, cli->name, &cli->base);
> > > + mutex_unlock(&drm->client_mutex);
> > > if (ret) {
> > > NV_PRINTK(err, cli, "Client allocation failed: %d\n",
> > > ret);
> > > goto done;
> > > @@ -602,7 +598,6 @@ nouveau_drm_device_fini(struct nouveau_drm *drm)
> > > mutex_unlock(&drm->clients_lock);
> > > nouveau_cli_fini(&drm->client);
> > > - nouveau_cli_fini(&drm->master);
> > > destroy_workqueue(drm->sched_wq);
> > > mutex_destroy(&drm->clients_lock);
> > > }
> > > @@ -618,13 +613,9 @@ nouveau_drm_device_init(struct nouveau_drm *drm)
> > > if (!drm->sched_wq)
> > > return -ENOMEM;
> > > - ret = nouveau_cli_init(drm, "DRM-master", &drm->master);
> > > - if (ret)
> > > - goto fail_wq;
> > > -
> > > ret = nouveau_cli_init(drm, "DRM", &drm->client);
> > > if (ret)
> > > - goto fail_master;
> > > + goto fail_wq;
> > > INIT_LIST_HEAD(&drm->clients);
> > > mutex_init(&drm->clients_lock);
> > > @@ -691,8 +682,6 @@ nouveau_drm_device_init(struct nouveau_drm *drm)
> > > fail_ttm:
> > > nouveau_vga_fini(drm);
> > > nouveau_cli_fini(&drm->client);
> > > -fail_master:
> > > - nouveau_cli_fini(&drm->master);
> > > fail_wq:
> > > destroy_workqueue(drm->sched_wq);
> > > return ret;
> > > @@ -706,9 +695,10 @@ nouveau_drm_device_del(struct nouveau_drm *drm)
> > > nvif_mmu_dtor(&drm->mmu);
> > > nvif_device_dtor(&drm->device);
> > > - nvif_client_dtor(&drm->master.base);
> > > + nvif_client_dtor(&drm->_client);
> > > nvif_parent_dtor(&drm->parent);
> > > + mutex_destroy(&drm->client_mutex);
> > > kfree(drm);
> > > }
> > > @@ -733,14 +723,15 @@ nouveau_drm_device_new(const struct drm_driver
> > > *drm_driver, struct device *paren
> > > drm->nvkm = device;
> > > nvif_parent_ctor(&nouveau_parent, &drm->parent);
> > > - drm->master.base.object.parent = &drm->parent;
> > > + mutex_init(&drm->client_mutex);
> > > + drm->_client.object.parent = &drm->parent;
> > > ret = nvif_driver_init(NULL, nouveau_config, nouveau_debug,
> > > "drm",
> > > - nouveau_name(parent), &drm->master.base);
> > > + nouveau_name(parent), &drm->_client);
> > > if (ret)
> > > goto done;
> > > - ret = nvif_device_ctor(&drm->master.base, "drmDevice", &drm->device);
> > > + ret = nvif_device_ctor(&drm->_client, "drmDevice", &drm->device);
> > > if (ret) {
> > > NV_ERROR(drm, "Device allocation failed: %d\n", ret);
> > > goto done;
> > > @@ -966,7 +957,7 @@ nouveau_do_suspend(struct nouveau_drm *drm, bool
> > > runtime)
> > > }
> > > NV_DEBUG(drm, "suspending object tree...\n");
> > > - ret = nvif_client_suspend(&drm->master.base);
> > > + ret = nvif_client_suspend(&drm->_client);
> > > if (ret)
> > > goto fail_client;
> > > @@ -991,7 +982,7 @@ nouveau_do_resume(struct nouveau_drm *drm, bool
> > > runtime)
> > > int ret = 0;
> > > NV_DEBUG(drm, "resuming object tree...\n");
> > > - ret = nvif_client_resume(&drm->master.base);
> > > + ret = nvif_client_resume(&drm->_client);
> > > if (ret) {
> > > NV_ERROR(drm, "Client resume failed with error: %d\n",
> > > ret);
> > > return ret;
> > > diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h
> > > b/drivers/gpu/drm/nouveau/nouveau_drv.h
> > > index 2535a50b99f3..630463668a76 100644
> > > --- a/drivers/gpu/drm/nouveau/nouveau_drv.h
> > > +++ b/drivers/gpu/drm/nouveau/nouveau_drv.h
> > > @@ -203,10 +203,11 @@ u_memcpya(uint64_t user, unsigned int nmemb,
> > > unsigned int size)
> > > struct nouveau_drm {
> > > struct nvkm_device *nvkm;
> > > struct nvif_parent parent;
> > > + struct mutex client_mutex;
> > > + struct nvif_client _client;
> > > struct nvif_device device;
> > > struct nvif_mmu mmu;
> > > - struct nouveau_cli master;
> > > struct nouveau_cli client;
> > > struct drm_device *dev;
> > > diff --git a/drivers/gpu/drm/nouveau/nouveau_mem.c
> > > b/drivers/gpu/drm/nouveau/nouveau_mem.c
> > > index b112b62dca3c..fac92fdbf9cc 100644
> > > --- a/drivers/gpu/drm/nouveau/nouveau_mem.c
> > > +++ b/drivers/gpu/drm/nouveau/nouveau_mem.c
> > > @@ -80,9 +80,9 @@ nouveau_mem_fini(struct nouveau_mem *mem)
> > > {
> > > nvif_vmm_put(&mem->drm->client.vmm.vmm, &mem->vma[1]);
> > > nvif_vmm_put(&mem->drm->client.vmm.vmm, &mem->vma[0]);
> > > - mutex_lock(&mem->drm->master.lock);
> > > + mutex_lock(&mem->drm->client_mutex);
> > > nvif_mem_dtor(&mem->mem);
> > > - mutex_unlock(&mem->drm->master.lock);
> > > + mutex_unlock(&mem->drm->client_mutex);
> > > }
> > > int
> > > @@ -113,11 +113,11 @@ nouveau_mem_host(struct ttm_resource *reg, struct
> > > ttm_tt *tt)
> > > else
> > > args.dma = tt->dma_address;
> > > - mutex_lock(&drm->master.lock);
> > > + mutex_lock(&drm->client_mutex);
> > > ret = nvif_mem_ctor_type(mmu, "ttmHostMem", mmu->mem, type,
> > > PAGE_SHIFT,
> > > reg->size,
> > > &args, sizeof(args), &mem->mem);
> > > - mutex_unlock(&drm->master.lock);
> > > + mutex_unlock(&drm->client_mutex);
> > > return ret;
> > > }
> > > @@ -130,7 +130,7 @@ nouveau_mem_vram(struct ttm_resource *reg, bool
> > > contig, u8 page)
> > > u64 size = ALIGN(reg->size, 1 << page);
> > > int ret;
> > > - mutex_lock(&drm->master.lock);
> > > + mutex_lock(&drm->client_mutex);
> > > switch (mmu->mem) {
> > > case NVIF_CLASS_MEM_GF100:
> > > ret = nvif_mem_ctor_type(mmu, "ttmVram", mmu->mem,
> > > @@ -154,7 +154,7 @@ nouveau_mem_vram(struct ttm_resource *reg, bool
> > > contig, u8 page)
> > > WARN_ON(1);
> > > break;
> > > }
> > > - mutex_unlock(&drm->master.lock);
> > > + mutex_unlock(&drm->client_mutex);
> > > reg->start = mem->mem.addr >> PAGE_SHIFT;
> > > return ret;
> > > --
> > > 2.45.1
> > >
>
