Check whether the starting address in the fault-around handler or the
faulty address in the fault handler is part of a huge page in order to
attempt a PMD sized PFN insertion into the VMA.

On builds with CONFIG_TRANSPARENT_HUGEPAGE enabled, if the mmap() user
address is PMD size aligned, if the GEM object is backed by shmem
buffers on mountpoints setting the 'huge=' option and if the shmem
backing store manages to allocate a huge folio, CPU mapping would then
benefit from significantly increased memcpy() performance. When these
conditions are met on a system with 2 MiB huge pages, an aligned copy
of 2 MiB would raise a single page fault instead of 4096.

v4:
- implement map_pages instead of huge_fault

Signed-off-by: Loïc Molinari <loic.molinari@collabora.com>
---
 drivers/gpu/drm/drm_gem_shmem_helper.c | 30 ++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)

diff --git a/drivers/gpu/drm/drm_gem_shmem_helper.c b/drivers/gpu/drm/drm_gem_shmem_helper.c
index 2a9fbc9c3712..b7238448d4f9 100644
--- a/drivers/gpu/drm/drm_gem_shmem_helper.c
+++ b/drivers/gpu/drm/drm_gem_shmem_helper.c
@@ -580,6 +580,26 @@ static bool drm_gem_shmem_fault_is_valid(struct drm_gem_object *obj,
 	return true;
 }
 
+static bool drm_gem_shmem_map_pmd(struct vm_fault *vmf, unsigned long addr,
+				  struct page *page)
+{
+#ifdef CONFIG_ARCH_SUPPORTS_PMD_PFNMAP
+	unsigned long pfn = page_to_pfn(page);
+	unsigned long paddr = pfn << PAGE_SHIFT;
+	bool aligned = (addr & ~PMD_MASK) == (paddr & ~PMD_MASK);
+
+	if (aligned &&
+	    pmd_none(*vmf->pmd) &&
+	    folio_test_pmd_mappable(page_folio(page))) {
+		pfn &= PMD_MASK >> PAGE_SHIFT;
+		if (vmf_insert_pfn_pmd(vmf, pfn, false) == VM_FAULT_NOPAGE)
+			return true;
+	}
+#endif
+
+	return false;
+}
+
 static vm_fault_t drm_gem_shmem_map_pages(struct vm_fault *vmf,
 					  pgoff_t start_pgoff,
 					  pgoff_t end_pgoff)
@@ -606,6 +626,11 @@ static vm_fault_t drm_gem_shmem_map_pages(struct vm_fault *vmf,
 		goto out;
 	}
 
+	if (drm_gem_shmem_map_pmd(vmf, addr, pages[start_pgoff])) {
+		ret = VM_FAULT_NOPAGE;
+		goto out;
+	}
+
 	/* Map a range of pages around the faulty address. */
 	do {
 		pfn = page_to_pfn(pages[start_pgoff]);
@@ -639,6 +664,11 @@ static vm_fault_t drm_gem_shmem_fault(struct vm_fault *vmf)
 		goto out;
 	}
 
+	if (drm_gem_shmem_map_pmd(vmf, vmf->address, pages[page_offset])) {
+		ret = VM_FAULT_NOPAGE;
+		goto out;
+	}
+
 	pfn = page_to_pfn(pages[page_offset]);
 	ret = vmf_insert_pfn(vma, vmf->address, pfn);
 
-- 
2.47.3