linux_dsm_epyc7002/drivers/gpu/drm/msm/msm_gem.c

1168 lines
27 KiB
C
Raw Normal View History

/*
* Copyright (C) 2013 Red Hat
* Author: Rob Clark <robdclark@gmail.com>
*
* This program is free software; you can redistribute it and/or modify it
* under the terms of the GNU General Public License version 2 as published by
* the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful, but WITHOUT
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
* more details.
*
* You should have received a copy of the GNU General Public License along with
* this program. If not, see <http://www.gnu.org/licenses/>.
*/
#include <linux/spinlock.h>
#include <linux/shmem_fs.h>
#include <linux/dma-buf.h>
#include <linux/pfn_t.h>
#include "msm_drv.h"
#include "msm_fence.h"
#include "msm_gem.h"
#include "msm_gpu.h"
#include "msm_mmu.h"
static void msm_gem_vunmap_locked(struct drm_gem_object *obj);
static dma_addr_t physaddr(struct drm_gem_object *obj)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
struct msm_drm_private *priv = obj->dev->dev_private;
return (((dma_addr_t)msm_obj->vram_node->start) << PAGE_SHIFT) +
priv->vram.paddr;
}
static bool use_pages(struct drm_gem_object *obj)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
return !msm_obj->vram_node;
}
/* allocate pages from VRAM carveout, used when no IOMMU: */
static struct page **get_pages_vram(struct drm_gem_object *obj, int npages)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
struct msm_drm_private *priv = obj->dev->dev_private;
dma_addr_t paddr;
struct page **p;
int ret, i;
p = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL);
if (!p)
return ERR_PTR(-ENOMEM);
spin_lock(&priv->vram.lock);
drm: Improve drm_mm search (and fix topdown allocation) with rbtrees The drm_mm range manager claimed to support top-down insertion, but it was neither searching for the top-most hole that could fit the allocation request nor fitting the request to the hole correctly. In order to search the range efficiently, we create a secondary index for the holes using either their size or their address. This index allows us to find the smallest hole or the hole at the bottom or top of the range efficiently, whilst keeping the hole stack to rapidly service evictions. v2: Search for holes both high and low. Rename flags to mode. v3: Discover rb_entry_safe() and use it! v4: Kerneldoc for enum drm_mm_insert_mode. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Cc: Alex Deucher <alexander.deucher@amd.com> Cc: "Christian König" <christian.koenig@amd.com> Cc: David Airlie <airlied@linux.ie> Cc: Russell King <rmk+kernel@armlinux.org.uk> Cc: Daniel Vetter <daniel.vetter@intel.com> Cc: Jani Nikula <jani.nikula@linux.intel.com> Cc: Sean Paul <seanpaul@chromium.org> Cc: Lucas Stach <l.stach@pengutronix.de> Cc: Christian Gmeiner <christian.gmeiner@gmail.com> Cc: Rob Clark <robdclark@gmail.com> Cc: Thierry Reding <thierry.reding@gmail.com> Cc: Stephen Warren <swarren@wwwdotorg.org> Cc: Alexandre Courbot <gnurou@gmail.com> Cc: Eric Anholt <eric@anholt.net> Cc: Sinclair Yeh <syeh@vmware.com> Cc: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Sinclair Yeh <syeh@vmware.com> # vmwgfx Reviewed-by: Lucas Stach <l.stach@pengutronix.de> #etnaviv Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: http://patchwork.freedesktop.org/patch/msgid/20170202210438.28702-1-chris@chris-wilson.co.uk
2017-02-03 04:04:38 +07:00
ret = drm_mm_insert_node(&priv->vram.mm, msm_obj->vram_node, npages);
spin_unlock(&priv->vram.lock);
if (ret) {
kvfree(p);
return ERR_PTR(ret);
}
paddr = physaddr(obj);
for (i = 0; i < npages; i++) {
p[i] = phys_to_page(paddr);
paddr += PAGE_SIZE;
}
return p;
}
static struct page **get_pages(struct drm_gem_object *obj)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
if (!msm_obj->pages) {
struct drm_device *dev = obj->dev;
struct page **p;
int npages = obj->size >> PAGE_SHIFT;
if (use_pages(obj))
p = drm_gem_get_pages(obj);
else
p = get_pages_vram(obj, npages);
if (IS_ERR(p)) {
DRM_DEV_ERROR(dev->dev, "could not get pages: %ld\n",
PTR_ERR(p));
return p;
}
msm_obj->pages = p;
msm_obj->sgt = drm_prime_pages_to_sg(p, npages);
if (IS_ERR(msm_obj->sgt)) {
void *ptr = ERR_CAST(msm_obj->sgt);
DRM_DEV_ERROR(dev->dev, "failed to allocate sgt\n");
msm_obj->sgt = NULL;
return ptr;
}
/* For non-cached buffers, ensure the new pages are clean
* because display controller, GPU, etc. are not coherent:
*/
if (msm_obj->flags & (MSM_BO_WC|MSM_BO_UNCACHED))
dma_map_sg(dev->dev, msm_obj->sgt->sgl,
msm_obj->sgt->nents, DMA_BIDIRECTIONAL);
}
return msm_obj->pages;
}
static void put_pages_vram(struct drm_gem_object *obj)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
struct msm_drm_private *priv = obj->dev->dev_private;
spin_lock(&priv->vram.lock);
drm_mm_remove_node(msm_obj->vram_node);
spin_unlock(&priv->vram.lock);
kvfree(msm_obj->pages);
}
static void put_pages(struct drm_gem_object *obj)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
if (msm_obj->pages) {
if (msm_obj->sgt) {
/* For non-cached buffers, ensure the new
* pages are clean because display controller,
* GPU, etc. are not coherent:
*/
if (msm_obj->flags & (MSM_BO_WC|MSM_BO_UNCACHED))
dma_unmap_sg(obj->dev->dev, msm_obj->sgt->sgl,
msm_obj->sgt->nents,
DMA_BIDIRECTIONAL);
sg_free_table(msm_obj->sgt);
kfree(msm_obj->sgt);
}
if (use_pages(obj))
drm_gem_put_pages(obj, msm_obj->pages, true, false);
else
put_pages_vram(obj);
msm_obj->pages = NULL;
}
}
struct page **msm_gem_get_pages(struct drm_gem_object *obj)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
struct page **p;
mutex_lock(&msm_obj->lock);
if (WARN_ON(msm_obj->madv != MSM_MADV_WILLNEED)) {
mutex_unlock(&msm_obj->lock);
return ERR_PTR(-EBUSY);
}
p = get_pages(obj);
mutex_unlock(&msm_obj->lock);
return p;
}
void msm_gem_put_pages(struct drm_gem_object *obj)
{
/* when we start tracking the pin count, then do something here */
}
int msm_gem_mmap_obj(struct drm_gem_object *obj,
struct vm_area_struct *vma)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
vma->vm_flags &= ~VM_PFNMAP;
vma->vm_flags |= VM_MIXEDMAP;
if (msm_obj->flags & MSM_BO_WC) {
vma->vm_page_prot = pgprot_writecombine(vm_get_page_prot(vma->vm_flags));
} else if (msm_obj->flags & MSM_BO_UNCACHED) {
vma->vm_page_prot = pgprot_noncached(vm_get_page_prot(vma->vm_flags));
} else {
/*
* Shunt off cached objs to shmem file so they have their own
* address_space (so unmap_mapping_range does what we want,
* in particular in the case of mmap'd dmabufs)
*/
fput(vma->vm_file);
get_file(obj->filp);
vma->vm_pgoff = 0;
vma->vm_file = obj->filp;
vma->vm_page_prot = vm_get_page_prot(vma->vm_flags);
}
return 0;
}
int msm_gem_mmap(struct file *filp, struct vm_area_struct *vma)
{
int ret;
ret = drm_gem_mmap(filp, vma);
if (ret) {
DBG("mmap failed: %d", ret);
return ret;
}
return msm_gem_mmap_obj(vma->vm_private_data, vma);
}
vm_fault_t msm_gem_fault(struct vm_fault *vmf)
{
struct vm_area_struct *vma = vmf->vma;
struct drm_gem_object *obj = vma->vm_private_data;
struct msm_gem_object *msm_obj = to_msm_bo(obj);
struct page **pages;
unsigned long pfn;
pgoff_t pgoff;
int err;
vm_fault_t ret;
/*
* vm_ops.open/drm_gem_mmap_obj and close get and put
* a reference on obj. So, we dont need to hold one here.
*/
err = mutex_lock_interruptible(&msm_obj->lock);
if (err) {
ret = VM_FAULT_NOPAGE;
goto out;
}
if (WARN_ON(msm_obj->madv != MSM_MADV_WILLNEED)) {
mutex_unlock(&msm_obj->lock);
return VM_FAULT_SIGBUS;
}
/* make sure we have pages attached now */
pages = get_pages(obj);
if (IS_ERR(pages)) {
ret = vmf_error(PTR_ERR(pages));
goto out_unlock;
}
/* We don't use vmf->pgoff since that has the fake offset: */
pgoff = (vmf->address - vma->vm_start) >> PAGE_SHIFT;
pfn = page_to_pfn(pages[pgoff]);
VERB("Inserting %p pfn %lx, pa %lx", (void *)vmf->address,
pfn, pfn << PAGE_SHIFT);
ret = vmf_insert_mixed(vma, vmf->address, __pfn_to_pfn_t(pfn, PFN_DEV));
out_unlock:
mutex_unlock(&msm_obj->lock);
out:
return ret;
}
/** get mmap offset */
static uint64_t mmap_offset(struct drm_gem_object *obj)
{
struct drm_device *dev = obj->dev;
struct msm_gem_object *msm_obj = to_msm_bo(obj);
int ret;
WARN_ON(!mutex_is_locked(&msm_obj->lock));
/* Make it mmapable */
ret = drm_gem_create_mmap_offset(obj);
if (ret) {
DRM_DEV_ERROR(dev->dev, "could not allocate mmap offset\n");
return 0;
}
return drm_vma_node_offset_addr(&obj->vma_node);
}
uint64_t msm_gem_mmap_offset(struct drm_gem_object *obj)
{
uint64_t offset;
struct msm_gem_object *msm_obj = to_msm_bo(obj);
mutex_lock(&msm_obj->lock);
offset = mmap_offset(obj);
mutex_unlock(&msm_obj->lock);
return offset;
}
static struct msm_gem_vma *add_vma(struct drm_gem_object *obj,
struct msm_gem_address_space *aspace)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
struct msm_gem_vma *vma;
WARN_ON(!mutex_is_locked(&msm_obj->lock));
vma = kzalloc(sizeof(*vma), GFP_KERNEL);
if (!vma)
return ERR_PTR(-ENOMEM);
vma->aspace = aspace;
list_add_tail(&vma->list, &msm_obj->vmas);
return vma;
}
static struct msm_gem_vma *lookup_vma(struct drm_gem_object *obj,
struct msm_gem_address_space *aspace)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
struct msm_gem_vma *vma;
WARN_ON(!mutex_is_locked(&msm_obj->lock));
list_for_each_entry(vma, &msm_obj->vmas, list) {
if (vma->aspace == aspace)
return vma;
}
return NULL;
}
static void del_vma(struct msm_gem_vma *vma)
{
if (!vma)
return;
list_del(&vma->list);
kfree(vma);
}
/* Called with msm_obj->lock locked */
static void
put_iova(struct drm_gem_object *obj)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
struct msm_gem_vma *vma, *tmp;
WARN_ON(!mutex_is_locked(&msm_obj->lock));
list_for_each_entry_safe(vma, tmp, &msm_obj->vmas, list) {
msm_gem_purge_vma(vma->aspace, vma);
msm_gem_close_vma(vma->aspace, vma);
del_vma(vma);
}
}
static int msm_gem_get_iova_locked(struct drm_gem_object *obj,
struct msm_gem_address_space *aspace, uint64_t *iova)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
struct msm_gem_vma *vma;
int ret = 0;
WARN_ON(!mutex_is_locked(&msm_obj->lock));
vma = lookup_vma(obj, aspace);
if (!vma) {
vma = add_vma(obj, aspace);
if (IS_ERR(vma))
return PTR_ERR(vma);
ret = msm_gem_init_vma(aspace, vma, obj->size >> PAGE_SHIFT);
if (ret) {
del_vma(vma);
return ret;
}
}
*iova = vma->iova;
return 0;
}
static int msm_gem_pin_iova(struct drm_gem_object *obj,
struct msm_gem_address_space *aspace)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
struct msm_gem_vma *vma;
struct page **pages;
int prot = IOMMU_READ;
if (!(msm_obj->flags & MSM_BO_GPU_READONLY))
prot |= IOMMU_WRITE;
WARN_ON(!mutex_is_locked(&msm_obj->lock));
if (WARN_ON(msm_obj->madv != MSM_MADV_WILLNEED))
return -EBUSY;
vma = lookup_vma(obj, aspace);
if (WARN_ON(!vma))
return -EINVAL;
pages = get_pages(obj);
if (IS_ERR(pages))
return PTR_ERR(pages);
return msm_gem_map_vma(aspace, vma, prot,
msm_obj->sgt, obj->size >> PAGE_SHIFT);
}
/* get iova and pin it. Should have a matching put */
int msm_gem_get_and_pin_iova(struct drm_gem_object *obj,
struct msm_gem_address_space *aspace, uint64_t *iova)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
u64 local;
int ret;
mutex_lock(&msm_obj->lock);
ret = msm_gem_get_iova_locked(obj, aspace, &local);
if (!ret)
ret = msm_gem_pin_iova(obj, aspace);
if (!ret)
*iova = local;
mutex_unlock(&msm_obj->lock);
return ret;
}
/*
* Get an iova but don't pin it. Doesn't need a put because iovas are currently
* valid for the life of the object
*/
int msm_gem_get_iova(struct drm_gem_object *obj,
struct msm_gem_address_space *aspace, uint64_t *iova)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
int ret;
mutex_lock(&msm_obj->lock);
ret = msm_gem_get_iova_locked(obj, aspace, iova);
mutex_unlock(&msm_obj->lock);
return ret;
}
/* get iova without taking a reference, used in places where you have
* already done a 'msm_gem_get_and_pin_iova' or 'msm_gem_get_iova'
*/
uint64_t msm_gem_iova(struct drm_gem_object *obj,
struct msm_gem_address_space *aspace)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
struct msm_gem_vma *vma;
mutex_lock(&msm_obj->lock);
vma = lookup_vma(obj, aspace);
mutex_unlock(&msm_obj->lock);
WARN_ON(!vma);
return vma ? vma->iova : 0;
}
/*
* Unpin a iova by updating the reference counts. The memory isn't actually
* purged until something else (shrinker, mm_notifier, destroy, etc) decides
* to get rid of it
*/
void msm_gem_unpin_iova(struct drm_gem_object *obj,
struct msm_gem_address_space *aspace)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
struct msm_gem_vma *vma;
mutex_lock(&msm_obj->lock);
vma = lookup_vma(obj, aspace);
if (!WARN_ON(!vma))
msm_gem_unmap_vma(aspace, vma);
mutex_unlock(&msm_obj->lock);
}
int msm_gem_dumb_create(struct drm_file *file, struct drm_device *dev,
struct drm_mode_create_dumb *args)
{
args->pitch = align_pitch(args->width, args->bpp);
args->size = PAGE_ALIGN(args->pitch * args->height);
return msm_gem_new_handle(dev, file, args->size,
MSM_BO_SCANOUT | MSM_BO_WC, &args->handle, "dumb");
}
int msm_gem_dumb_map_offset(struct drm_file *file, struct drm_device *dev,
uint32_t handle, uint64_t *offset)
{
struct drm_gem_object *obj;
int ret = 0;
/* GEM does all our handle to object mapping */
obj = drm_gem_object_lookup(file, handle);
if (obj == NULL) {
ret = -ENOENT;
goto fail;
}
*offset = msm_gem_mmap_offset(obj);
drm_gem_object_put_unlocked(obj);
fail:
return ret;
}
static void *get_vaddr(struct drm_gem_object *obj, unsigned madv)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
int ret = 0;
mutex_lock(&msm_obj->lock);
if (WARN_ON(msm_obj->madv > madv)) {
DRM_DEV_ERROR(obj->dev->dev, "Invalid madv state: %u vs %u\n",
msm_obj->madv, madv);
mutex_unlock(&msm_obj->lock);
return ERR_PTR(-EBUSY);
}
/* increment vmap_count *before* vmap() call, so shrinker can
* check vmap_count (is_vunmapable()) outside of msm_obj->lock.
* This guarantees that we won't try to msm_gem_vunmap() this
* same object from within the vmap() call (while we already
* hold msm_obj->lock)
*/
msm_obj->vmap_count++;
if (!msm_obj->vaddr) {
struct page **pages = get_pages(obj);
if (IS_ERR(pages)) {
ret = PTR_ERR(pages);
goto fail;
}
msm_obj->vaddr = vmap(pages, obj->size >> PAGE_SHIFT,
VM_MAP, pgprot_writecombine(PAGE_KERNEL));
if (msm_obj->vaddr == NULL) {
ret = -ENOMEM;
goto fail;
}
}
mutex_unlock(&msm_obj->lock);
return msm_obj->vaddr;
fail:
msm_obj->vmap_count--;
mutex_unlock(&msm_obj->lock);
return ERR_PTR(ret);
}
void *msm_gem_get_vaddr(struct drm_gem_object *obj)
{
return get_vaddr(obj, MSM_MADV_WILLNEED);
}
/*
* Don't use this! It is for the very special case of dumping
* submits from GPU hangs or faults, were the bo may already
* be MSM_MADV_DONTNEED, but we know the buffer is still on the
* active list.
*/
void *msm_gem_get_vaddr_active(struct drm_gem_object *obj)
{
return get_vaddr(obj, __MSM_MADV_PURGED);
}
void msm_gem_put_vaddr(struct drm_gem_object *obj)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
mutex_lock(&msm_obj->lock);
WARN_ON(msm_obj->vmap_count < 1);
msm_obj->vmap_count--;
mutex_unlock(&msm_obj->lock);
}
/* Update madvise status, returns true if not purged, else
* false or -errno.
*/
int msm_gem_madvise(struct drm_gem_object *obj, unsigned madv)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
mutex_lock(&msm_obj->lock);
WARN_ON(!mutex_is_locked(&obj->dev->struct_mutex));
if (msm_obj->madv != __MSM_MADV_PURGED)
msm_obj->madv = madv;
madv = msm_obj->madv;
mutex_unlock(&msm_obj->lock);
return (madv != __MSM_MADV_PURGED);
}
void msm_gem_purge(struct drm_gem_object *obj, enum msm_gem_lock subclass)
{
struct drm_device *dev = obj->dev;
struct msm_gem_object *msm_obj = to_msm_bo(obj);
WARN_ON(!mutex_is_locked(&dev->struct_mutex));
WARN_ON(!is_purgeable(msm_obj));
WARN_ON(obj->import_attach);
mutex_lock_nested(&msm_obj->lock, subclass);
put_iova(obj);
msm_gem_vunmap_locked(obj);
put_pages(obj);
msm_obj->madv = __MSM_MADV_PURGED;
drm_vma_node_unmap(&obj->vma_node, dev->anon_inode->i_mapping);
drm_gem_free_mmap_offset(obj);
/* Our goal here is to return as much of the memory as
* is possible back to the system as we are called from OOM.
* To do this we must instruct the shmfs to drop all of its
* backing pages, *now*.
*/
shmem_truncate_range(file_inode(obj->filp), 0, (loff_t)-1);
invalidate_mapping_pages(file_inode(obj->filp)->i_mapping,
0, (loff_t)-1);
mutex_unlock(&msm_obj->lock);
}
static void msm_gem_vunmap_locked(struct drm_gem_object *obj)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
WARN_ON(!mutex_is_locked(&msm_obj->lock));
if (!msm_obj->vaddr || WARN_ON(!is_vunmapable(msm_obj)))
return;
vunmap(msm_obj->vaddr);
msm_obj->vaddr = NULL;
}
void msm_gem_vunmap(struct drm_gem_object *obj, enum msm_gem_lock subclass)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
mutex_lock_nested(&msm_obj->lock, subclass);
msm_gem_vunmap_locked(obj);
mutex_unlock(&msm_obj->lock);
}
/* must be called before _move_to_active().. */
int msm_gem_sync_object(struct drm_gem_object *obj,
struct msm_fence_context *fctx, bool exclusive)
{
struct reservation_object_list *fobj;
dma-buf: Rename struct fence to dma_fence I plan to usurp the short name of struct fence for a core kernel struct, and so I need to rename the specialised fence/timeline for DMA operations to make room. A consensus was reached in https://lists.freedesktop.org/archives/dri-devel/2016-July/113083.html that making clear this fence applies to DMA operations was a good thing. Since then the patch has grown a bit as usage increases, so hopefully it remains a good thing! (v2...: rebase, rerun spatch) v3: Compile on msm, spotted a manual fixup that I broke. v4: Try again for msm, sorry Daniel coccinelle script: @@ @@ - struct fence + struct dma_fence @@ @@ - struct fence_ops + struct dma_fence_ops @@ @@ - struct fence_cb + struct dma_fence_cb @@ @@ - struct fence_array + struct dma_fence_array @@ @@ - enum fence_flag_bits + enum dma_fence_flag_bits @@ @@ ( - fence_init + dma_fence_init | - fence_release + dma_fence_release | - fence_free + dma_fence_free | - fence_get + dma_fence_get | - fence_get_rcu + dma_fence_get_rcu | - fence_put + dma_fence_put | - fence_signal + dma_fence_signal | - fence_signal_locked + dma_fence_signal_locked | - fence_default_wait + dma_fence_default_wait | - fence_add_callback + dma_fence_add_callback | - fence_remove_callback + dma_fence_remove_callback | - fence_enable_sw_signaling + dma_fence_enable_sw_signaling | - fence_is_signaled_locked + dma_fence_is_signaled_locked | - fence_is_signaled + dma_fence_is_signaled | - fence_is_later + dma_fence_is_later | - fence_later + dma_fence_later | - fence_wait_timeout + dma_fence_wait_timeout | - fence_wait_any_timeout + dma_fence_wait_any_timeout | - fence_wait + dma_fence_wait | - fence_context_alloc + dma_fence_context_alloc | - fence_array_create + dma_fence_array_create | - to_fence_array + to_dma_fence_array | - fence_is_array + dma_fence_is_array | - trace_fence_emit + trace_dma_fence_emit | - FENCE_TRACE + DMA_FENCE_TRACE | - FENCE_WARN + DMA_FENCE_WARN | - FENCE_ERR + DMA_FENCE_ERR ) ( ... ) Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk> Acked-by: Sumit Semwal <sumit.semwal@linaro.org> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: http://patchwork.freedesktop.org/patch/msgid/20161025120045.28839-1-chris@chris-wilson.co.uk
2016-10-25 19:00:45 +07:00
struct dma_fence *fence;
int i, ret;
fobj = reservation_object_get_list(obj->resv);
if (!fobj || (fobj->shared_count == 0)) {
fence = reservation_object_get_excl(obj->resv);
/* don't need to wait on our own fences, since ring is fifo */
if (fence && (fence->context != fctx->context)) {
dma-buf: Rename struct fence to dma_fence I plan to usurp the short name of struct fence for a core kernel struct, and so I need to rename the specialised fence/timeline for DMA operations to make room. A consensus was reached in https://lists.freedesktop.org/archives/dri-devel/2016-July/113083.html that making clear this fence applies to DMA operations was a good thing. Since then the patch has grown a bit as usage increases, so hopefully it remains a good thing! (v2...: rebase, rerun spatch) v3: Compile on msm, spotted a manual fixup that I broke. v4: Try again for msm, sorry Daniel coccinelle script: @@ @@ - struct fence + struct dma_fence @@ @@ - struct fence_ops + struct dma_fence_ops @@ @@ - struct fence_cb + struct dma_fence_cb @@ @@ - struct fence_array + struct dma_fence_array @@ @@ - enum fence_flag_bits + enum dma_fence_flag_bits @@ @@ ( - fence_init + dma_fence_init | - fence_release + dma_fence_release | - fence_free + dma_fence_free | - fence_get + dma_fence_get | - fence_get_rcu + dma_fence_get_rcu | - fence_put + dma_fence_put | - fence_signal + dma_fence_signal | - fence_signal_locked + dma_fence_signal_locked | - fence_default_wait + dma_fence_default_wait | - fence_add_callback + dma_fence_add_callback | - fence_remove_callback + dma_fence_remove_callback | - fence_enable_sw_signaling + dma_fence_enable_sw_signaling | - fence_is_signaled_locked + dma_fence_is_signaled_locked | - fence_is_signaled + dma_fence_is_signaled | - fence_is_later + dma_fence_is_later | - fence_later + dma_fence_later | - fence_wait_timeout + dma_fence_wait_timeout | - fence_wait_any_timeout + dma_fence_wait_any_timeout | - fence_wait + dma_fence_wait | - fence_context_alloc + dma_fence_context_alloc | - fence_array_create + dma_fence_array_create | - to_fence_array + to_dma_fence_array | - fence_is_array + dma_fence_is_array | - trace_fence_emit + trace_dma_fence_emit | - FENCE_TRACE + DMA_FENCE_TRACE | - FENCE_WARN + DMA_FENCE_WARN | - FENCE_ERR + DMA_FENCE_ERR ) ( ... ) Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk> Acked-by: Sumit Semwal <sumit.semwal@linaro.org> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: http://patchwork.freedesktop.org/patch/msgid/20161025120045.28839-1-chris@chris-wilson.co.uk
2016-10-25 19:00:45 +07:00
ret = dma_fence_wait(fence, true);
if (ret)
return ret;
}
}
if (!exclusive || !fobj)
return 0;
for (i = 0; i < fobj->shared_count; i++) {
fence = rcu_dereference_protected(fobj->shared[i],
reservation_object_held(obj->resv));
if (fence->context != fctx->context) {
dma-buf: Rename struct fence to dma_fence I plan to usurp the short name of struct fence for a core kernel struct, and so I need to rename the specialised fence/timeline for DMA operations to make room. A consensus was reached in https://lists.freedesktop.org/archives/dri-devel/2016-July/113083.html that making clear this fence applies to DMA operations was a good thing. Since then the patch has grown a bit as usage increases, so hopefully it remains a good thing! (v2...: rebase, rerun spatch) v3: Compile on msm, spotted a manual fixup that I broke. v4: Try again for msm, sorry Daniel coccinelle script: @@ @@ - struct fence + struct dma_fence @@ @@ - struct fence_ops + struct dma_fence_ops @@ @@ - struct fence_cb + struct dma_fence_cb @@ @@ - struct fence_array + struct dma_fence_array @@ @@ - enum fence_flag_bits + enum dma_fence_flag_bits @@ @@ ( - fence_init + dma_fence_init | - fence_release + dma_fence_release | - fence_free + dma_fence_free | - fence_get + dma_fence_get | - fence_get_rcu + dma_fence_get_rcu | - fence_put + dma_fence_put | - fence_signal + dma_fence_signal | - fence_signal_locked + dma_fence_signal_locked | - fence_default_wait + dma_fence_default_wait | - fence_add_callback + dma_fence_add_callback | - fence_remove_callback + dma_fence_remove_callback | - fence_enable_sw_signaling + dma_fence_enable_sw_signaling | - fence_is_signaled_locked + dma_fence_is_signaled_locked | - fence_is_signaled + dma_fence_is_signaled | - fence_is_later + dma_fence_is_later | - fence_later + dma_fence_later | - fence_wait_timeout + dma_fence_wait_timeout | - fence_wait_any_timeout + dma_fence_wait_any_timeout | - fence_wait + dma_fence_wait | - fence_context_alloc + dma_fence_context_alloc | - fence_array_create + dma_fence_array_create | - to_fence_array + to_dma_fence_array | - fence_is_array + dma_fence_is_array | - trace_fence_emit + trace_dma_fence_emit | - FENCE_TRACE + DMA_FENCE_TRACE | - FENCE_WARN + DMA_FENCE_WARN | - FENCE_ERR + DMA_FENCE_ERR ) ( ... ) Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk> Acked-by: Sumit Semwal <sumit.semwal@linaro.org> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: http://patchwork.freedesktop.org/patch/msgid/20161025120045.28839-1-chris@chris-wilson.co.uk
2016-10-25 19:00:45 +07:00
ret = dma_fence_wait(fence, true);
if (ret)
return ret;
}
}
return 0;
}
void msm_gem_move_to_active(struct drm_gem_object *obj,
dma-buf: Rename struct fence to dma_fence I plan to usurp the short name of struct fence for a core kernel struct, and so I need to rename the specialised fence/timeline for DMA operations to make room. A consensus was reached in https://lists.freedesktop.org/archives/dri-devel/2016-July/113083.html that making clear this fence applies to DMA operations was a good thing. Since then the patch has grown a bit as usage increases, so hopefully it remains a good thing! (v2...: rebase, rerun spatch) v3: Compile on msm, spotted a manual fixup that I broke. v4: Try again for msm, sorry Daniel coccinelle script: @@ @@ - struct fence + struct dma_fence @@ @@ - struct fence_ops + struct dma_fence_ops @@ @@ - struct fence_cb + struct dma_fence_cb @@ @@ - struct fence_array + struct dma_fence_array @@ @@ - enum fence_flag_bits + enum dma_fence_flag_bits @@ @@ ( - fence_init + dma_fence_init | - fence_release + dma_fence_release | - fence_free + dma_fence_free | - fence_get + dma_fence_get | - fence_get_rcu + dma_fence_get_rcu | - fence_put + dma_fence_put | - fence_signal + dma_fence_signal | - fence_signal_locked + dma_fence_signal_locked | - fence_default_wait + dma_fence_default_wait | - fence_add_callback + dma_fence_add_callback | - fence_remove_callback + dma_fence_remove_callback | - fence_enable_sw_signaling + dma_fence_enable_sw_signaling | - fence_is_signaled_locked + dma_fence_is_signaled_locked | - fence_is_signaled + dma_fence_is_signaled | - fence_is_later + dma_fence_is_later | - fence_later + dma_fence_later | - fence_wait_timeout + dma_fence_wait_timeout | - fence_wait_any_timeout + dma_fence_wait_any_timeout | - fence_wait + dma_fence_wait | - fence_context_alloc + dma_fence_context_alloc | - fence_array_create + dma_fence_array_create | - to_fence_array + to_dma_fence_array | - fence_is_array + dma_fence_is_array | - trace_fence_emit + trace_dma_fence_emit | - FENCE_TRACE + DMA_FENCE_TRACE | - FENCE_WARN + DMA_FENCE_WARN | - FENCE_ERR + DMA_FENCE_ERR ) ( ... ) Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk> Acked-by: Sumit Semwal <sumit.semwal@linaro.org> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: http://patchwork.freedesktop.org/patch/msgid/20161025120045.28839-1-chris@chris-wilson.co.uk
2016-10-25 19:00:45 +07:00
struct msm_gpu *gpu, bool exclusive, struct dma_fence *fence)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
WARN_ON(msm_obj->madv != MSM_MADV_WILLNEED);
msm_obj->gpu = gpu;
if (exclusive)
reservation_object_add_excl_fence(obj->resv, fence);
else
reservation_object_add_shared_fence(obj->resv, fence);
list_del_init(&msm_obj->mm_list);
list_add_tail(&msm_obj->mm_list, &gpu->active_list);
}
void msm_gem_move_to_inactive(struct drm_gem_object *obj)
{
struct drm_device *dev = obj->dev;
struct msm_drm_private *priv = dev->dev_private;
struct msm_gem_object *msm_obj = to_msm_bo(obj);
WARN_ON(!mutex_is_locked(&dev->struct_mutex));
msm_obj->gpu = NULL;
list_del_init(&msm_obj->mm_list);
list_add_tail(&msm_obj->mm_list, &priv->inactive_list);
}
int msm_gem_cpu_prep(struct drm_gem_object *obj, uint32_t op, ktime_t *timeout)
{
bool write = !!(op & MSM_PREP_WRITE);
unsigned long remain =
op & MSM_PREP_NOSYNC ? 0 : timeout_to_jiffies(timeout);
long ret;
ret = reservation_object_wait_timeout_rcu(obj->resv, write,
true, remain);
if (ret == 0)
return remain == 0 ? -EBUSY : -ETIMEDOUT;
else if (ret < 0)
return ret;
/* TODO cache maintenance */
return 0;
}
int msm_gem_cpu_fini(struct drm_gem_object *obj)
{
/* TODO cache maintenance */
return 0;
}
#ifdef CONFIG_DEBUG_FS
dma-buf: Rename struct fence to dma_fence I plan to usurp the short name of struct fence for a core kernel struct, and so I need to rename the specialised fence/timeline for DMA operations to make room. A consensus was reached in https://lists.freedesktop.org/archives/dri-devel/2016-July/113083.html that making clear this fence applies to DMA operations was a good thing. Since then the patch has grown a bit as usage increases, so hopefully it remains a good thing! (v2...: rebase, rerun spatch) v3: Compile on msm, spotted a manual fixup that I broke. v4: Try again for msm, sorry Daniel coccinelle script: @@ @@ - struct fence + struct dma_fence @@ @@ - struct fence_ops + struct dma_fence_ops @@ @@ - struct fence_cb + struct dma_fence_cb @@ @@ - struct fence_array + struct dma_fence_array @@ @@ - enum fence_flag_bits + enum dma_fence_flag_bits @@ @@ ( - fence_init + dma_fence_init | - fence_release + dma_fence_release | - fence_free + dma_fence_free | - fence_get + dma_fence_get | - fence_get_rcu + dma_fence_get_rcu | - fence_put + dma_fence_put | - fence_signal + dma_fence_signal | - fence_signal_locked + dma_fence_signal_locked | - fence_default_wait + dma_fence_default_wait | - fence_add_callback + dma_fence_add_callback | - fence_remove_callback + dma_fence_remove_callback | - fence_enable_sw_signaling + dma_fence_enable_sw_signaling | - fence_is_signaled_locked + dma_fence_is_signaled_locked | - fence_is_signaled + dma_fence_is_signaled | - fence_is_later + dma_fence_is_later | - fence_later + dma_fence_later | - fence_wait_timeout + dma_fence_wait_timeout | - fence_wait_any_timeout + dma_fence_wait_any_timeout | - fence_wait + dma_fence_wait | - fence_context_alloc + dma_fence_context_alloc | - fence_array_create + dma_fence_array_create | - to_fence_array + to_dma_fence_array | - fence_is_array + dma_fence_is_array | - trace_fence_emit + trace_dma_fence_emit | - FENCE_TRACE + DMA_FENCE_TRACE | - FENCE_WARN + DMA_FENCE_WARN | - FENCE_ERR + DMA_FENCE_ERR ) ( ... ) Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk> Acked-by: Sumit Semwal <sumit.semwal@linaro.org> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: http://patchwork.freedesktop.org/patch/msgid/20161025120045.28839-1-chris@chris-wilson.co.uk
2016-10-25 19:00:45 +07:00
static void describe_fence(struct dma_fence *fence, const char *type,
struct seq_file *m)
{
dma-buf: Rename struct fence to dma_fence I plan to usurp the short name of struct fence for a core kernel struct, and so I need to rename the specialised fence/timeline for DMA operations to make room. A consensus was reached in https://lists.freedesktop.org/archives/dri-devel/2016-July/113083.html that making clear this fence applies to DMA operations was a good thing. Since then the patch has grown a bit as usage increases, so hopefully it remains a good thing! (v2...: rebase, rerun spatch) v3: Compile on msm, spotted a manual fixup that I broke. v4: Try again for msm, sorry Daniel coccinelle script: @@ @@ - struct fence + struct dma_fence @@ @@ - struct fence_ops + struct dma_fence_ops @@ @@ - struct fence_cb + struct dma_fence_cb @@ @@ - struct fence_array + struct dma_fence_array @@ @@ - enum fence_flag_bits + enum dma_fence_flag_bits @@ @@ ( - fence_init + dma_fence_init | - fence_release + dma_fence_release | - fence_free + dma_fence_free | - fence_get + dma_fence_get | - fence_get_rcu + dma_fence_get_rcu | - fence_put + dma_fence_put | - fence_signal + dma_fence_signal | - fence_signal_locked + dma_fence_signal_locked | - fence_default_wait + dma_fence_default_wait | - fence_add_callback + dma_fence_add_callback | - fence_remove_callback + dma_fence_remove_callback | - fence_enable_sw_signaling + dma_fence_enable_sw_signaling | - fence_is_signaled_locked + dma_fence_is_signaled_locked | - fence_is_signaled + dma_fence_is_signaled | - fence_is_later + dma_fence_is_later | - fence_later + dma_fence_later | - fence_wait_timeout + dma_fence_wait_timeout | - fence_wait_any_timeout + dma_fence_wait_any_timeout | - fence_wait + dma_fence_wait | - fence_context_alloc + dma_fence_context_alloc | - fence_array_create + dma_fence_array_create | - to_fence_array + to_dma_fence_array | - fence_is_array + dma_fence_is_array | - trace_fence_emit + trace_dma_fence_emit | - FENCE_TRACE + DMA_FENCE_TRACE | - FENCE_WARN + DMA_FENCE_WARN | - FENCE_ERR + DMA_FENCE_ERR ) ( ... ) Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk> Acked-by: Sumit Semwal <sumit.semwal@linaro.org> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: http://patchwork.freedesktop.org/patch/msgid/20161025120045.28839-1-chris@chris-wilson.co.uk
2016-10-25 19:00:45 +07:00
if (!dma_fence_is_signaled(fence))
seq_printf(m, "\t%9s: %s %s seq %llu\n", type,
fence->ops->get_driver_name(fence),
fence->ops->get_timeline_name(fence),
fence->seqno);
}
void msm_gem_describe(struct drm_gem_object *obj, struct seq_file *m)
{
struct msm_gem_object *msm_obj = to_msm_bo(obj);
struct reservation_object *robj = obj->resv;
struct reservation_object_list *fobj;
dma-buf: Rename struct fence to dma_fence I plan to usurp the short name of struct fence for a core kernel struct, and so I need to rename the specialised fence/timeline for DMA operations to make room. A consensus was reached in https://lists.freedesktop.org/archives/dri-devel/2016-July/113083.html that making clear this fence applies to DMA operations was a good thing. Since then the patch has grown a bit as usage increases, so hopefully it remains a good thing! (v2...: rebase, rerun spatch) v3: Compile on msm, spotted a manual fixup that I broke. v4: Try again for msm, sorry Daniel coccinelle script: @@ @@ - struct fence + struct dma_fence @@ @@ - struct fence_ops + struct dma_fence_ops @@ @@ - struct fence_cb + struct dma_fence_cb @@ @@ - struct fence_array + struct dma_fence_array @@ @@ - enum fence_flag_bits + enum dma_fence_flag_bits @@ @@ ( - fence_init + dma_fence_init | - fence_release + dma_fence_release | - fence_free + dma_fence_free | - fence_get + dma_fence_get | - fence_get_rcu + dma_fence_get_rcu | - fence_put + dma_fence_put | - fence_signal + dma_fence_signal | - fence_signal_locked + dma_fence_signal_locked | - fence_default_wait + dma_fence_default_wait | - fence_add_callback + dma_fence_add_callback | - fence_remove_callback + dma_fence_remove_callback | - fence_enable_sw_signaling + dma_fence_enable_sw_signaling | - fence_is_signaled_locked + dma_fence_is_signaled_locked | - fence_is_signaled + dma_fence_is_signaled | - fence_is_later + dma_fence_is_later | - fence_later + dma_fence_later | - fence_wait_timeout + dma_fence_wait_timeout | - fence_wait_any_timeout + dma_fence_wait_any_timeout | - fence_wait + dma_fence_wait | - fence_context_alloc + dma_fence_context_alloc | - fence_array_create + dma_fence_array_create | - to_fence_array + to_dma_fence_array | - fence_is_array + dma_fence_is_array | - trace_fence_emit + trace_dma_fence_emit | - FENCE_TRACE + DMA_FENCE_TRACE | - FENCE_WARN + DMA_FENCE_WARN | - FENCE_ERR + DMA_FENCE_ERR ) ( ... ) Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk> Acked-by: Sumit Semwal <sumit.semwal@linaro.org> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: http://patchwork.freedesktop.org/patch/msgid/20161025120045.28839-1-chris@chris-wilson.co.uk
2016-10-25 19:00:45 +07:00
struct dma_fence *fence;
struct msm_gem_vma *vma;
uint64_t off = drm_vma_node_start(&obj->vma_node);
const char *madv;
mutex_lock(&msm_obj->lock);
switch (msm_obj->madv) {
case __MSM_MADV_PURGED:
madv = " purged";
break;
case MSM_MADV_DONTNEED:
madv = " purgeable";
break;
case MSM_MADV_WILLNEED:
default:
madv = "";
break;
}
seq_printf(m, "%08x: %c %2d (%2d) %08llx %p",
msm_obj->flags, is_active(msm_obj) ? 'A' : 'I',
obj->name, kref_read(&obj->refcount),
off, msm_obj->vaddr);
seq_printf(m, " %08zu %9s %-32s\n", obj->size, madv, msm_obj->name);
if (!list_empty(&msm_obj->vmas)) {
seq_puts(m, " vmas:");
list_for_each_entry(vma, &msm_obj->vmas, list)
seq_printf(m, " [%s: %08llx,%s,inuse=%d]", vma->aspace->name,
vma->iova, vma->mapped ? "mapped" : "unmapped",
vma->inuse);
seq_puts(m, "\n");
}
rcu_read_lock();
fobj = rcu_dereference(robj->fence);
if (fobj) {
unsigned int i, shared_count = fobj->shared_count;
for (i = 0; i < shared_count; i++) {
fence = rcu_dereference(fobj->shared[i]);
describe_fence(fence, "Shared", m);
}
}
fence = rcu_dereference(robj->fence_excl);
if (fence)
describe_fence(fence, "Exclusive", m);
rcu_read_unlock();
mutex_unlock(&msm_obj->lock);
}
void msm_gem_describe_objects(struct list_head *list, struct seq_file *m)
{
struct msm_gem_object *msm_obj;
int count = 0;
size_t size = 0;
seq_puts(m, " flags id ref offset kaddr size madv name\n");
list_for_each_entry(msm_obj, list, mm_list) {
struct drm_gem_object *obj = &msm_obj->base;
seq_puts(m, " ");
msm_gem_describe(obj, m);
count++;
size += obj->size;
}
seq_printf(m, "Total %d objects, %zu bytes\n", count, size);
}
#endif
/* don't call directly! Use drm_gem_object_put() and friends */
void msm_gem_free_object(struct drm_gem_object *obj)
{
struct drm_device *dev = obj->dev;
struct msm_gem_object *msm_obj = to_msm_bo(obj);
WARN_ON(!mutex_is_locked(&dev->struct_mutex));
/* object should not be on active list: */
WARN_ON(is_active(msm_obj));
list_del(&msm_obj->mm_list);
mutex_lock(&msm_obj->lock);
put_iova(obj);
if (obj->import_attach) {
if (msm_obj->vaddr)
dma_buf_vunmap(obj->import_attach->dmabuf, msm_obj->vaddr);
/* Don't drop the pages for imported dmabuf, as they are not
* ours, just free the array we allocated:
*/
if (msm_obj->pages)
kvfree(msm_obj->pages);
drm_prime_gem_destroy(obj, msm_obj->sgt);
} else {
msm_gem_vunmap_locked(obj);
put_pages(obj);
}
drm_gem_object_release(obj);
mutex_unlock(&msm_obj->lock);
kfree(msm_obj);
}
/* convenience method to construct a GEM buffer object, and userspace handle */
int msm_gem_new_handle(struct drm_device *dev, struct drm_file *file,
uint32_t size, uint32_t flags, uint32_t *handle,
char *name)
{
struct drm_gem_object *obj;
int ret;
obj = msm_gem_new(dev, size, flags);
if (IS_ERR(obj))
return PTR_ERR(obj);
if (name)
msm_gem_object_set_name(obj, "%s", name);
ret = drm_gem_handle_create(file, obj, handle);
/* drop reference from allocate - handle holds it now */
drm_gem_object_put_unlocked(obj);
return ret;
}
static int msm_gem_new_impl(struct drm_device *dev,
uint32_t size, uint32_t flags,
struct reservation_object *resv,
struct drm_gem_object **obj,
bool struct_mutex_locked)
{
struct msm_drm_private *priv = dev->dev_private;
struct msm_gem_object *msm_obj;
switch (flags & MSM_BO_CACHE_MASK) {
case MSM_BO_UNCACHED:
case MSM_BO_CACHED:
case MSM_BO_WC:
break;
default:
DRM_DEV_ERROR(dev->dev, "invalid cache flag: %x\n",
(flags & MSM_BO_CACHE_MASK));
return -EINVAL;
}
msm_obj = kzalloc(sizeof(*msm_obj), GFP_KERNEL);
if (!msm_obj)
return -ENOMEM;
mutex_init(&msm_obj->lock);
msm_obj->flags = flags;
msm_obj->madv = MSM_MADV_WILLNEED;
if (resv)
msm_obj->base.resv = resv;
INIT_LIST_HEAD(&msm_obj->submit_entry);
INIT_LIST_HEAD(&msm_obj->vmas);
if (struct_mutex_locked) {
WARN_ON(!mutex_is_locked(&dev->struct_mutex));
list_add_tail(&msm_obj->mm_list, &priv->inactive_list);
} else {
mutex_lock(&dev->struct_mutex);
list_add_tail(&msm_obj->mm_list, &priv->inactive_list);
mutex_unlock(&dev->struct_mutex);
}
*obj = &msm_obj->base;
return 0;
}
static struct drm_gem_object *_msm_gem_new(struct drm_device *dev,
uint32_t size, uint32_t flags, bool struct_mutex_locked)
{
struct msm_drm_private *priv = dev->dev_private;
struct drm_gem_object *obj = NULL;
bool use_vram = false;
int ret;
size = PAGE_ALIGN(size);
if (!msm_use_mmu(dev))
use_vram = true;
else if ((flags & (MSM_BO_STOLEN | MSM_BO_SCANOUT)) && priv->vram.size)
use_vram = true;
if (WARN_ON(use_vram && !priv->vram.size))
return ERR_PTR(-EINVAL);
/* Disallow zero sized objects as they make the underlying
* infrastructure grumpy
*/
if (size == 0)
return ERR_PTR(-EINVAL);
ret = msm_gem_new_impl(dev, size, flags, NULL, &obj, struct_mutex_locked);
if (ret)
goto fail;
if (use_vram) {
struct msm_gem_vma *vma;
struct page **pages;
drm/msm: fix WARN_ON in add_vma() with no iommu While I was testing the upcoming adv7533 CEC support with my Dragonboard c410 I encountered this warning several times during boot: [ 4.408309] WARNING: CPU: 3 PID: 1347 at drivers/gpu/drm/msm/msm_gem.c:312 add_vma+0x78/0x88 [msm] [ 4.412951] Modules linked in: snd_soc_hdmi_codec adv7511 cec qcom_wcnss_pil msm mdt_loader drm_kms_helper msm_rng rng_core drm [ 4.421728] CPU: 3 PID: 1347 Comm: kworker/3:3 Not tainted 4.13.0-rc1-dragonboard #111 [ 4.433090] Hardware name: Qualcomm Technologies, Inc. APQ 8016 SBC (DT) [ 4.441081] Workqueue: events deferred_probe_work_func [ 4.447929] task: ffff800031243600 task.stack: ffff800003394000 [ 4.453023] PC is at add_vma+0x78/0x88 [msm] [ 4.458823] LR is at _msm_gem_new+0xd4/0x188 [msm] [ 4.463207] pc : [<ffff000000ac01f8>] lr : [<ffff000000ac06b4>] pstate: 40000145 [ 4.467811] sp : ffff8000033978a0 [ 4.475357] x29: ffff8000033978a0 x28: ffff8000031dea18 [ 4.478572] x27: ffff800003933a00 x26: ffff800003b39800 [ 4.483953] x25: ffff8000338ff800 x24: 0000000000000001 [ 4.489249] x23: 0000000000000000 x22: ffff800003b39800 [ 4.494544] x21: ffff8000338ff800 x20: 0000000000000000 [ 4.499839] x19: ffff800003932600 x18: 0000000000000001 [ 4.505135] x17: 0000ffff8969e9e0 x16: ffff7e00000ce7a0 [ 4.510429] x15: ffffffffffffffff x14: ffff8000833977ef [ 4.515724] x13: ffff8000033977f3 x12: 0000000000000038 [ 4.521020] x11: 0101010101010101 x10: ffffff7f7fff7f7f [ 4.526315] x9 : 0000000000000000 x8 : ffff800003932800 [ 4.531633] x7 : 0000000000000000 x6 : 000000000000003f [ 4.531644] x5 : 0000000000000040 x4 : 0000000000000000 [ 4.531650] x3 : ffff800031243600 x2 : 0000000000000000 [ 4.531655] x1 : 0000000000000000 x0 : 0000000000000000 [ 4.531670] Call trace: [ 4.531676] Exception stack(0xffff8000033976c0 to 0xffff8000033977f0) [ 4.531683] 76c0: ffff800003932600 0001000000000000 ffff8000033978a0 ffff000000ac01f8 [ 4.531688] 76e0: 0000000000000140 0000000000000000 ffff800003932550 ffff800003397780 [ 4.531694] 7700: ffff800003397730 ffff000008261ce8 0000000000000000 ffff8000031d2f80 [ 4.531699] 7720: ffff800003397800 ffff0000081d671c 0000000000000140 0000000000000000 [ 4.531705] 7740: ffff000000ac04c0 0000000000004003 ffff800003397908 00000000014080c0 [ 4.531710] 7760: 0000000000000000 ffff800003b39800 0000000000000000 0000000000000000 [ 4.531716] 7780: 0000000000000000 ffff800031243600 0000000000000000 0000000000000040 [ 4.531721] 77a0: 000000000000003f 0000000000000000 ffff800003932800 0000000000000000 [ 4.531726] 77c0: ffffff7f7fff7f7f 0101010101010101 0000000000000038 ffff8000033977f3 [ 4.531730] 77e0: ffff8000833977ef ffffffffffffffff [ 4.531881] [<ffff000000ac01f8>] add_vma+0x78/0x88 [msm] [ 4.532011] [<ffff000000ac06b4>] _msm_gem_new+0xd4/0x188 [msm] [ 4.532134] [<ffff000000ac1900>] msm_gem_new+0x10/0x18 [msm] [ 4.532260] [<ffff000000acb274>] msm_dsi_host_modeset_init+0x17c/0x268 [msm] [ 4.532384] [<ffff000000ac9024>] msm_dsi_modeset_init+0x34/0x1b8 [msm] [ 4.532504] [<ffff000000ab6168>] modeset_init+0x408/0x488 [msm] [ 4.532623] [<ffff000000ab6c4c>] mdp5_kms_init+0x2b4/0x338 [msm] [ 4.532745] [<ffff000000abeff8>] msm_drm_bind+0x218/0x4e8 [msm] [ 4.532755] [<ffff00000855d744>] try_to_bring_up_master+0x1f4/0x318 [ 4.532762] [<ffff00000855d900>] component_add+0x98/0x180 [ 4.532887] [<ffff000000ac8da0>] dsi_dev_probe+0x18/0x28 [msm] [ 4.532895] [<ffff000008565fe8>] platform_drv_probe+0x58/0xc0 [ 4.532901] [<ffff00000856410c>] driver_probe_device+0x324/0x458 [ 4.532907] [<ffff00000856440c>] __device_attach_driver+0xac/0x170 [ 4.532913] [<ffff000008561ef4>] bus_for_each_drv+0x4c/0x98 [ 4.532918] [<ffff000008563c38>] __device_attach+0xc0/0x160 [ 4.532924] [<ffff000008564530>] device_initial_probe+0x10/0x18 [ 4.532929] [<ffff000008562f84>] bus_probe_device+0x94/0xa0 [ 4.532934] [<ffff0000085635d4>] deferred_probe_work_func+0x8c/0xe8 [ 4.532941] [<ffff0000080d79bc>] process_one_work+0x1d4/0x330 [ 4.532946] [<ffff0000080d7b60>] worker_thread+0x48/0x468 [ 4.532952] [<ffff0000080ddae4>] kthread+0x12c/0x130 [ 4.532958] [<ffff000008082f10>] ret_from_fork+0x10/0x40 [ 4.532962] ---[ end trace b1ac6888ec40b0bb ]--- Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Rob Clark <robdclark@gmail.com>
2017-07-30 19:42:36 +07:00
struct msm_gem_object *msm_obj = to_msm_bo(obj);
mutex_lock(&msm_obj->lock);
vma = add_vma(obj, NULL);
drm/msm: fix WARN_ON in add_vma() with no iommu While I was testing the upcoming adv7533 CEC support with my Dragonboard c410 I encountered this warning several times during boot: [ 4.408309] WARNING: CPU: 3 PID: 1347 at drivers/gpu/drm/msm/msm_gem.c:312 add_vma+0x78/0x88 [msm] [ 4.412951] Modules linked in: snd_soc_hdmi_codec adv7511 cec qcom_wcnss_pil msm mdt_loader drm_kms_helper msm_rng rng_core drm [ 4.421728] CPU: 3 PID: 1347 Comm: kworker/3:3 Not tainted 4.13.0-rc1-dragonboard #111 [ 4.433090] Hardware name: Qualcomm Technologies, Inc. APQ 8016 SBC (DT) [ 4.441081] Workqueue: events deferred_probe_work_func [ 4.447929] task: ffff800031243600 task.stack: ffff800003394000 [ 4.453023] PC is at add_vma+0x78/0x88 [msm] [ 4.458823] LR is at _msm_gem_new+0xd4/0x188 [msm] [ 4.463207] pc : [<ffff000000ac01f8>] lr : [<ffff000000ac06b4>] pstate: 40000145 [ 4.467811] sp : ffff8000033978a0 [ 4.475357] x29: ffff8000033978a0 x28: ffff8000031dea18 [ 4.478572] x27: ffff800003933a00 x26: ffff800003b39800 [ 4.483953] x25: ffff8000338ff800 x24: 0000000000000001 [ 4.489249] x23: 0000000000000000 x22: ffff800003b39800 [ 4.494544] x21: ffff8000338ff800 x20: 0000000000000000 [ 4.499839] x19: ffff800003932600 x18: 0000000000000001 [ 4.505135] x17: 0000ffff8969e9e0 x16: ffff7e00000ce7a0 [ 4.510429] x15: ffffffffffffffff x14: ffff8000833977ef [ 4.515724] x13: ffff8000033977f3 x12: 0000000000000038 [ 4.521020] x11: 0101010101010101 x10: ffffff7f7fff7f7f [ 4.526315] x9 : 0000000000000000 x8 : ffff800003932800 [ 4.531633] x7 : 0000000000000000 x6 : 000000000000003f [ 4.531644] x5 : 0000000000000040 x4 : 0000000000000000 [ 4.531650] x3 : ffff800031243600 x2 : 0000000000000000 [ 4.531655] x1 : 0000000000000000 x0 : 0000000000000000 [ 4.531670] Call trace: [ 4.531676] Exception stack(0xffff8000033976c0 to 0xffff8000033977f0) [ 4.531683] 76c0: ffff800003932600 0001000000000000 ffff8000033978a0 ffff000000ac01f8 [ 4.531688] 76e0: 0000000000000140 0000000000000000 ffff800003932550 ffff800003397780 [ 4.531694] 7700: ffff800003397730 ffff000008261ce8 0000000000000000 ffff8000031d2f80 [ 4.531699] 7720: ffff800003397800 ffff0000081d671c 0000000000000140 0000000000000000 [ 4.531705] 7740: ffff000000ac04c0 0000000000004003 ffff800003397908 00000000014080c0 [ 4.531710] 7760: 0000000000000000 ffff800003b39800 0000000000000000 0000000000000000 [ 4.531716] 7780: 0000000000000000 ffff800031243600 0000000000000000 0000000000000040 [ 4.531721] 77a0: 000000000000003f 0000000000000000 ffff800003932800 0000000000000000 [ 4.531726] 77c0: ffffff7f7fff7f7f 0101010101010101 0000000000000038 ffff8000033977f3 [ 4.531730] 77e0: ffff8000833977ef ffffffffffffffff [ 4.531881] [<ffff000000ac01f8>] add_vma+0x78/0x88 [msm] [ 4.532011] [<ffff000000ac06b4>] _msm_gem_new+0xd4/0x188 [msm] [ 4.532134] [<ffff000000ac1900>] msm_gem_new+0x10/0x18 [msm] [ 4.532260] [<ffff000000acb274>] msm_dsi_host_modeset_init+0x17c/0x268 [msm] [ 4.532384] [<ffff000000ac9024>] msm_dsi_modeset_init+0x34/0x1b8 [msm] [ 4.532504] [<ffff000000ab6168>] modeset_init+0x408/0x488 [msm] [ 4.532623] [<ffff000000ab6c4c>] mdp5_kms_init+0x2b4/0x338 [msm] [ 4.532745] [<ffff000000abeff8>] msm_drm_bind+0x218/0x4e8 [msm] [ 4.532755] [<ffff00000855d744>] try_to_bring_up_master+0x1f4/0x318 [ 4.532762] [<ffff00000855d900>] component_add+0x98/0x180 [ 4.532887] [<ffff000000ac8da0>] dsi_dev_probe+0x18/0x28 [msm] [ 4.532895] [<ffff000008565fe8>] platform_drv_probe+0x58/0xc0 [ 4.532901] [<ffff00000856410c>] driver_probe_device+0x324/0x458 [ 4.532907] [<ffff00000856440c>] __device_attach_driver+0xac/0x170 [ 4.532913] [<ffff000008561ef4>] bus_for_each_drv+0x4c/0x98 [ 4.532918] [<ffff000008563c38>] __device_attach+0xc0/0x160 [ 4.532924] [<ffff000008564530>] device_initial_probe+0x10/0x18 [ 4.532929] [<ffff000008562f84>] bus_probe_device+0x94/0xa0 [ 4.532934] [<ffff0000085635d4>] deferred_probe_work_func+0x8c/0xe8 [ 4.532941] [<ffff0000080d79bc>] process_one_work+0x1d4/0x330 [ 4.532946] [<ffff0000080d7b60>] worker_thread+0x48/0x468 [ 4.532952] [<ffff0000080ddae4>] kthread+0x12c/0x130 [ 4.532958] [<ffff000008082f10>] ret_from_fork+0x10/0x40 [ 4.532962] ---[ end trace b1ac6888ec40b0bb ]--- Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Rob Clark <robdclark@gmail.com>
2017-07-30 19:42:36 +07:00
mutex_unlock(&msm_obj->lock);
if (IS_ERR(vma)) {
ret = PTR_ERR(vma);
goto fail;
}
to_msm_bo(obj)->vram_node = &vma->node;
drm_gem_private_object_init(dev, obj, size);
pages = get_pages(obj);
if (IS_ERR(pages)) {
ret = PTR_ERR(pages);
goto fail;
}
vma->iova = physaddr(obj);
} else {
ret = drm_gem_object_init(dev, obj, size);
if (ret)
goto fail;
}
return obj;
fail:
drm_gem_object_put_unlocked(obj);
return ERR_PTR(ret);
}
struct drm_gem_object *msm_gem_new_locked(struct drm_device *dev,
uint32_t size, uint32_t flags)
{
return _msm_gem_new(dev, size, flags, true);
}
struct drm_gem_object *msm_gem_new(struct drm_device *dev,
uint32_t size, uint32_t flags)
{
return _msm_gem_new(dev, size, flags, false);
}
struct drm_gem_object *msm_gem_import(struct drm_device *dev,
struct dma_buf *dmabuf, struct sg_table *sgt)
{
struct msm_gem_object *msm_obj;
struct drm_gem_object *obj;
uint32_t size;
int ret, npages;
/* if we don't have IOMMU, don't bother pretending we can import: */
if (!msm_use_mmu(dev)) {
DRM_DEV_ERROR(dev->dev, "cannot import without IOMMU\n");
return ERR_PTR(-EINVAL);
}
size = PAGE_ALIGN(dmabuf->size);
ret = msm_gem_new_impl(dev, size, MSM_BO_WC, dmabuf->resv, &obj, false);
if (ret)
goto fail;
drm_gem_private_object_init(dev, obj, size);
npages = size / PAGE_SIZE;
msm_obj = to_msm_bo(obj);
mutex_lock(&msm_obj->lock);
msm_obj->sgt = sgt;
msm_obj->pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL);
if (!msm_obj->pages) {
mutex_unlock(&msm_obj->lock);
ret = -ENOMEM;
goto fail;
}
ret = drm_prime_sg_to_page_addr_arrays(sgt, msm_obj->pages, NULL, npages);
if (ret) {
mutex_unlock(&msm_obj->lock);
goto fail;
}
mutex_unlock(&msm_obj->lock);
return obj;
fail:
drm_gem_object_put_unlocked(obj);
return ERR_PTR(ret);
}
static void *_msm_gem_kernel_new(struct drm_device *dev, uint32_t size,
uint32_t flags, struct msm_gem_address_space *aspace,
struct drm_gem_object **bo, uint64_t *iova, bool locked)
{
void *vaddr;
struct drm_gem_object *obj = _msm_gem_new(dev, size, flags, locked);
int ret;
if (IS_ERR(obj))
return ERR_CAST(obj);
if (iova) {
ret = msm_gem_get_and_pin_iova(obj, aspace, iova);
if (ret)
goto err;
}
vaddr = msm_gem_get_vaddr(obj);
if (IS_ERR(vaddr)) {
msm_gem_unpin_iova(obj, aspace);
ret = PTR_ERR(vaddr);
goto err;
}
if (bo)
*bo = obj;
return vaddr;
err:
if (locked)
drm_gem_object_put(obj);
else
drm_gem_object_put_unlocked(obj);
return ERR_PTR(ret);
}
void *msm_gem_kernel_new(struct drm_device *dev, uint32_t size,
uint32_t flags, struct msm_gem_address_space *aspace,
struct drm_gem_object **bo, uint64_t *iova)
{
return _msm_gem_kernel_new(dev, size, flags, aspace, bo, iova, false);
}
void *msm_gem_kernel_new_locked(struct drm_device *dev, uint32_t size,
uint32_t flags, struct msm_gem_address_space *aspace,
struct drm_gem_object **bo, uint64_t *iova)
{
return _msm_gem_kernel_new(dev, size, flags, aspace, bo, iova, true);
}
void msm_gem_kernel_put(struct drm_gem_object *bo,
struct msm_gem_address_space *aspace, bool locked)
{
if (IS_ERR_OR_NULL(bo))
return;
msm_gem_put_vaddr(bo);
msm_gem_unpin_iova(bo, aspace);
if (locked)
drm_gem_object_put(bo);
else
drm_gem_object_put_unlocked(bo);
}
void msm_gem_object_set_name(struct drm_gem_object *bo, const char *fmt, ...)
{
struct msm_gem_object *msm_obj = to_msm_bo(bo);
va_list ap;
if (!fmt)
return;
va_start(ap, fmt);
vsnprintf(msm_obj->name, sizeof(msm_obj->name), fmt, ap);
va_end(ap);
}