p71924506
/
linux_old1
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

drm/nouveau/dmem: use dma addresses during migration copies 

Removes the need for temporary VMM mappings.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
This commit is contained in:
Ben Skeggs 2019-02-15 15:50:16 +10:00
parent fd5e985643
commit a788ade4f6
1 changed files with 53 additions and 106 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

159
drivers/gpu/drm/nouveau/nouveau_dmem.c
View File

@ -80,17 +80,11 @@ struct nouveau_dmem {
struct mutex mutex; struct mutex mutex;
}; };
struct nouveau_migrate_hmem {
struct scatterlist *sg;
struct nouveau_mem mem;
unsigned long npages;
struct nvif_vma vma;
};
struct nouveau_dmem_fault { struct nouveau_dmem_fault {
struct nouveau_drm *drm; struct nouveau_drm *drm;
struct nouveau_fence *fence; struct nouveau_fence *fence;
struct nouveau_migrate_hmem hmem; dma_addr_t *dma;
unsigned long npages;
}; };
struct nouveau_migrate { struct nouveau_migrate {
@ -98,87 +92,10 @@ struct nouveau_migrate {
struct nouveau_drm *drm; struct nouveau_drm *drm;
struct nouveau_fence *fence; struct nouveau_fence *fence;
unsigned long npages; unsigned long npages;
struct nouveau_migrate_hmem hmem; dma_addr_t *dma;
unsigned long dma_nr;
}; };
static void
nouveau_migrate_hmem_fini(struct nouveau_drm *drm,
struct nouveau_migrate_hmem *hmem)
{
struct nvif_vmm *vmm = &drm->client.vmm.vmm;
nouveau_mem_fini(&hmem->mem);
nvif_vmm_put(vmm, &hmem->vma);
if (hmem->sg) {
dma_unmap_sg_attrs(drm->dev->dev, hmem->sg,
hmem->npages, DMA_BIDIRECTIONAL,
DMA_ATTR_SKIP_CPU_SYNC);
kfree(hmem->sg);
hmem->sg = NULL;
}
}
static int
nouveau_migrate_hmem_init(struct nouveau_drm *drm,
struct nouveau_migrate_hmem *hmem,
unsigned long npages,
const unsigned long *pfns)
{
struct nvif_vmm *vmm = &drm->client.vmm.vmm;
unsigned long i;
int ret;
hmem->sg = kzalloc(npages * sizeof(*hmem->sg), GFP_KERNEL);
if (hmem->sg == NULL)
return -ENOMEM;
for (i = 0, hmem->npages = 0; hmem->npages < npages; ++i) {
struct page *page;
if (!pfns[i] || pfns[i] == MIGRATE_PFN_ERROR)
continue;
page = migrate_pfn_to_page(pfns[i]);
if (page == NULL) {
ret = -EINVAL;
goto error;
}
sg_set_page(&hmem->sg[hmem->npages], page, PAGE_SIZE, 0);
hmem->npages++;
}
sg_mark_end(&hmem->sg[hmem->npages - 1]);
i = dma_map_sg_attrs(drm->dev->dev, hmem->sg, hmem->npages,
DMA_BIDIRECTIONAL, DMA_ATTR_SKIP_CPU_SYNC);
if (i != hmem->npages) {
ret = -ENOMEM;
goto error;
}
ret = nouveau_mem_sgl(&hmem->mem, &drm->client,
hmem->npages, hmem->sg);
if (ret)
goto error;
ret = nvif_vmm_get(vmm, LAZY, false, hmem->mem.mem.page,
0, hmem->mem.mem.size, &hmem->vma);
if (ret)
goto error;
ret = nouveau_mem_map(&hmem->mem, vmm, &hmem->vma);
if (ret)
goto error;
return 0;
error:
nouveau_migrate_hmem_fini(drm, hmem);
return ret;
}
static void static void
nouveau_dmem_free(struct hmm_devmem *devmem, struct page *page) nouveau_dmem_free(struct hmm_devmem *devmem, struct page *page)
{ {
@ -218,7 +135,8 @@ nouveau_dmem_fault_alloc_and_copy(struct vm_area_struct *vma,
{ {
struct nouveau_dmem_fault *fault = private; struct nouveau_dmem_fault *fault = private;
struct nouveau_drm *drm = fault->drm; struct nouveau_drm *drm = fault->drm;
unsigned long addr, i, c, npages = 0; struct device *dev = drm->dev->dev;
unsigned long addr, i, npages = 0;
nouveau_migrate_copy_t copy; nouveau_migrate_copy_t copy;
int ret; int ret;
@ -243,14 +161,14 @@ nouveau_dmem_fault_alloc_and_copy(struct vm_area_struct *vma,
npages++; npages++;
} }
/* Create scatter list FIXME: get rid of scatter list */ /* Allocate storage for DMA addresses, so we can unmap later. */
ret = nouveau_migrate_hmem_init(drm, &fault->hmem, npages, dst_pfns); fault->dma = kmalloc(sizeof(*fault->dma) * npages, GFP_KERNEL);
if (ret) if (!fault->dma)
goto error; goto error;
/* Copy things over */ /* Copy things over */
copy = drm->dmem->migrate.copy_func; copy = drm->dmem->migrate.copy_func;
for (addr = start, i = c = 0; addr < end; addr += PAGE_SIZE, i++) { for (addr = start, i = 0; addr < end; addr += PAGE_SIZE, i++) {
struct nouveau_dmem_chunk *chunk; struct nouveau_dmem_chunk *chunk;
struct page *spage, *dpage; struct page *spage, *dpage;
u64 src_addr, dst_addr; u64 src_addr, dst_addr;
@ -259,9 +177,6 @@ nouveau_dmem_fault_alloc_and_copy(struct vm_area_struct *vma,
if (!dpage || dst_pfns[i] == MIGRATE_PFN_ERROR) if (!dpage || dst_pfns[i] == MIGRATE_PFN_ERROR)
continue; continue;
dst_addr = fault->hmem.vma.addr + (c << PAGE_SHIFT);
c++;
spage = migrate_pfn_to_page(src_pfns[i]); spage = migrate_pfn_to_page(src_pfns[i]);
if (!spage || !(src_pfns[i] & MIGRATE_PFN_MIGRATE)) { if (!spage || !(src_pfns[i] & MIGRATE_PFN_MIGRATE)) {
dst_pfns[i] = MIGRATE_PFN_ERROR; dst_pfns[i] = MIGRATE_PFN_ERROR;
@ -269,11 +184,23 @@ nouveau_dmem_fault_alloc_and_copy(struct vm_area_struct *vma,
continue; continue;
} }
fault->dma[fault->npages] =
dma_map_page_attrs(dev, dpage, 0, PAGE_SIZE,
PCI_DMA_BIDIRECTIONAL,
DMA_ATTR_SKIP_CPU_SYNC);
if (dma_mapping_error(dev, fault->dma[fault->npages])) {
dst_pfns[i] = MIGRATE_PFN_ERROR;
__free_page(dpage);
continue;
}
dst_addr = fault->dma[fault->npages++];
chunk = (void *)hmm_devmem_page_get_drvdata(spage); chunk = (void *)hmm_devmem_page_get_drvdata(spage);
src_addr = page_to_pfn(spage) - chunk->pfn_first; src_addr = page_to_pfn(spage) - chunk->pfn_first;
src_addr = (src_addr << PAGE_SHIFT) + chunk->bo->bo.offset; src_addr = (src_addr << PAGE_SHIFT) + chunk->bo->bo.offset;
ret = copy(drm, 1, NOUVEAU_APER_VIRT, dst_addr, ret = copy(drm, 1, NOUVEAU_APER_HOST, dst_addr,
NOUVEAU_APER_VRAM, src_addr); NOUVEAU_APER_VRAM, src_addr);
if (ret) { if (ret) {
dst_pfns[i] = MIGRATE_PFN_ERROR; dst_pfns[i] = MIGRATE_PFN_ERROR;
@ -321,7 +248,12 @@ void nouveau_dmem_fault_finalize_and_map(struct vm_area_struct *vma,
* the hmem object below (nouveau_migrate_hmem_fini()). * the hmem object below (nouveau_migrate_hmem_fini()).
*/ */
} }
nouveau_migrate_hmem_fini(drm, &fault->hmem);
while (fault->npages--) {
dma_unmap_page(drm->dev->dev, fault->dma[fault->npages],
PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
}
kfree(fault->dma);
} }
static const struct migrate_vma_ops nouveau_dmem_fault_migrate_ops = { static const struct migrate_vma_ops nouveau_dmem_fault_migrate_ops = {
@ -732,7 +664,8 @@ nouveau_dmem_migrate_alloc_and_copy(struct vm_area_struct *vma,
{ {
struct nouveau_migrate *migrate = private; struct nouveau_migrate *migrate = private;
struct nouveau_drm *drm = migrate->drm; struct nouveau_drm *drm = migrate->drm;
unsigned long addr, i, c, npages = 0; struct device *dev = drm->dev->dev;
unsigned long addr, i, npages = 0;
nouveau_migrate_copy_t copy; nouveau_migrate_copy_t copy;
int ret; int ret;
@ -758,14 +691,14 @@ nouveau_dmem_migrate_alloc_and_copy(struct vm_area_struct *vma,
if (!npages) if (!npages)
return; return;
/* Create scatter list FIXME: get rid of scatter list */ /* Allocate storage for DMA addresses, so we can unmap later. */
ret = nouveau_migrate_hmem_init(drm, &migrate->hmem, npages, src_pfns); migrate->dma = kmalloc(sizeof(*migrate->dma) * npages, GFP_KERNEL);
if (ret) if (!migrate->dma)
goto error; goto error;
/* Copy things over */ /* Copy things over */
copy = drm->dmem->migrate.copy_func; copy = drm->dmem->migrate.copy_func;
for (addr = start, i = c = 0; addr < end; addr += PAGE_SIZE, i++) { for (addr = start, i = 0; addr < end; addr += PAGE_SIZE, i++) {
struct nouveau_dmem_chunk *chunk; struct nouveau_dmem_chunk *chunk;
struct page *spage, *dpage; struct page *spage, *dpage;
u64 src_addr, dst_addr; u64 src_addr, dst_addr;
@ -785,11 +718,20 @@ nouveau_dmem_migrate_alloc_and_copy(struct vm_area_struct *vma,
continue; continue;
} }
src_addr = migrate->hmem.vma.addr + (c << PAGE_SHIFT); migrate->dma[migrate->dma_nr] =
c++; dma_map_page_attrs(dev, spage, 0, PAGE_SIZE,
PCI_DMA_BIDIRECTIONAL,
DMA_ATTR_SKIP_CPU_SYNC);
if (dma_mapping_error(dev, migrate->dma[migrate->dma_nr])) {
nouveau_dmem_page_free_locked(drm, dpage);
dst_pfns[i] = 0;
continue;
}
src_addr = migrate->dma[migrate->dma_nr++];
ret = copy(drm, 1, NOUVEAU_APER_VRAM, dst_addr, ret = copy(drm, 1, NOUVEAU_APER_VRAM, dst_addr,
NOUVEAU_APER_VIRT, src_addr); NOUVEAU_APER_HOST, src_addr);
if (ret) { if (ret) {
nouveau_dmem_page_free_locked(drm, dpage); nouveau_dmem_page_free_locked(drm, dpage);
dst_pfns[i] = 0; dst_pfns[i] = 0;
@ -836,7 +778,12 @@ void nouveau_dmem_migrate_finalize_and_map(struct vm_area_struct *vma,
* the hmem object below (nouveau_migrate_hmem_fini()) ? * the hmem object below (nouveau_migrate_hmem_fini()) ?
*/ */
} }
nouveau_migrate_hmem_fini(drm, &migrate->hmem);
while (migrate->dma_nr--) {
dma_unmap_page(drm->dev->dev, migrate->dma[migrate->dma_nr],
PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
}
kfree(migrate->dma);
/* /*
* FIXME optimization: update GPU page table to point to newly * FIXME optimization: update GPU page table to point to newly