rpms/kernel/devel linux-2.6.29-pat-fixes.patch, NONE, 1.1 TODO, 1.60, 1.61 kernel.spec, 1.1484, 1.1485

[Adam Jackson]

Author: ajax

Update of /cvs/pkgs/rpms/kernel/devel
In directory cvs1.fedora.phx.redhat.com:/tmp/cvs-serv18285

Modified Files:
	TODO kernel.spec 
Added Files:
	linux-2.6.29-pat-fixes.patch 
Log Message:
* Mon Mar 30 2009 Adam Jackson <ajax at redhat.com>
- linux-2.6.29-pat-fixes.patch: Fix PAT/GTT interaction


linux-2.6.29-pat-fixes.patch:

--- NEW FILE linux-2.6.29-pat-fixes.patch ---
>From 7806028c712e4947b764636f6668849edd3cae0b Mon Sep 17 00:00:00 2001
From: Adam Jackson <ajax at redhat.com>
Date: Thu, 26 Mar 2009 17:47:49 -0400
Subject: [PATCH] pat hacks

Required for GTT maps on intel to work.  See also:
http://bugs.freedesktop.org/show_bug.cgi?id=20803

Stolen from http://bugs.freedesktop.org/attachment.cgi?id=24172
with minor fixes.

---
 arch/x86/mm/pat.c   |   50 ++++++++++++++++++++++++++++++++++++++------------
 arch/x86/pci/i386.c |    3 +++
 include/linux/mm.h  |   15 +++++++++++++--
 mm/memory.c         |    6 ++++--
 4 files changed, 58 insertions(+), 16 deletions(-)

diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c
index e0ab173..026247e 100644
--- a/arch/x86/mm/pat.c
+++ b/arch/x86/mm/pat.c
@@ -365,11 +365,14 @@ int reserve_memtype(u64 start, u64 end, unsigned long req_type,
 		*new_type = actual_type;
 
 	is_range_ram = pat_pagerange_is_ram(start, end);
-	if (is_range_ram == 1)
+	if (is_range_ram == 1) {
+		printk(KERN_WARNING "reserve_memtype: calling reserve_ram_pages_type for 0x%llx 0x%llx %lu\n", start, end, req_type);
 		return reserve_ram_pages_type(start, end, req_type,
 					      new_type);
-	else if (is_range_ram < 0)
+	} else if (is_range_ram < 0) {
+		printk(KERN_WARNING "reserve_memtype: is_range_ram < 0 for 0x%llx 0x%llx %lu\n", start, end, req_type);
 		return -EINVAL;
+	}
 
 	new  = kmalloc(sizeof(struct memtype), GFP_KERNEL);
 	if (!new)
@@ -641,14 +644,21 @@ static int reserve_pfn_range(u64 paddr, unsigned long size, pgprot_t *vma_prot,
 	is_ram = pat_pagerange_is_ram(paddr, paddr + size);
 
 	/*
-	 * reserve_pfn_range() doesn't support RAM pages.
+	 * reserve_pfn_range() doesn't support RAM pages. Maintain the current
+	 * behavior with RAM pages by returning success.
 	 */
-	if (is_ram != 0)
-		return -EINVAL;
+	if (is_ram != 0) {
+		printk(KERN_WARNING "reserve_pfn_range: is_ram is %d for 0x%llx!\n",
+			is_ram, paddr);
+		return 0;
+	}
 
 	ret = reserve_memtype(paddr, paddr + size, want_flags, &flags);
-	if (ret)
+	if (ret) {
+		printk(KERN_WARNING "reserve_pfn_range: reserve_memtype ret %d for 0x%llx 0x%lx 0x%lx\n",
+			ret, paddr, size, want_flags);
 		return ret;
+	}
 
 	if (flags != want_flags) {
 		if (strict_prot || !is_new_memtype_allowed(want_flags, flags)) {
@@ -739,7 +749,12 @@ int track_pfn_vma_copy(struct vm_area_struct *vma)
 			return -EINVAL;
 		}
 		pgprot = __pgprot(prot);
-		return reserve_pfn_range(paddr, vma_size, &pgprot, 1);
+		retval = reserve_pfn_range(paddr, vma_size, &pgprot, 1);
+		if (retval) {
+			printk(KERN_WARNING "track_pfn_vma_copy linear - 0x%llx 0x%lx 0x%lx failed\n",
+				paddr, vma_size, prot);
+		}
+		return retval;
 	}
 
 	/* reserve entire vma page by page, using pfn and prot from pte */
@@ -749,8 +764,11 @@ int track_pfn_vma_copy(struct vm_area_struct *vma)
 
 		pgprot = __pgprot(prot);
 		retval = reserve_pfn_range(paddr, PAGE_SIZE, &pgprot, 1);
-		if (retval)
+		if (retval) {
+			printk(KERN_WARNING "track_pfn_vma_copy not linear - 0x%llx 0x%lx failed\n",
+				paddr, prot);
 			goto cleanup_ret;
+		}
 	}
 	return 0;
 
@@ -795,8 +813,13 @@ int track_pfn_vma_new(struct vm_area_struct *vma, pgprot_t *prot,
 
 	if (is_linear_pfn_mapping(vma)) {
 		/* reserve the whole chunk starting from vm_pgoff */
-		paddr = (resource_size_t)vma->vm_pgoff << PAGE_SHIFT;
-		return reserve_pfn_range(paddr, vma_size, prot, 0);
+		paddr = (resource_size_t)pfn << PAGE_SHIFT;
+		retval = reserve_pfn_range(paddr, vma_size, prot, 0);
+		if (retval) {
+			printk(KERN_WARNING "track_pfn_vma_new linear - 0x%llx 0x%lx 0x%x failed\n",
+				paddr, size, (int) pgprot_val(*prot));
+		}
+		return retval;
 	}
 
 	/* reserve page by page using pfn and size */
@@ -804,8 +827,11 @@ int track_pfn_vma_new(struct vm_area_struct *vma, pgprot_t *prot,
 	for (i = 0; i < size; i += PAGE_SIZE) {
 		paddr = base_paddr + i;
 		retval = reserve_pfn_range(paddr, PAGE_SIZE, prot, 0);
-		if (retval)
+		if (retval) {
+			printk(KERN_WARNING "track_pfn_vma_new not linear - 0x%llx 0x%x failed\n",
+				paddr, (int) pgprot_val(*prot));
 			goto cleanup_ret;
+		}
 	}
 	return 0;
 
@@ -816,7 +842,7 @@ cleanup_ret:
 		free_pfn_range(paddr, PAGE_SIZE);
 	}
 
-	return retval;
+	return 0;
 }
 
 /*
diff --git a/arch/x86/pci/i386.c b/arch/x86/pci/i386.c
index 5ead808..f234a37 100644
--- a/arch/x86/pci/i386.c
+++ b/arch/x86/pci/i386.c
@@ -319,6 +319,9 @@ int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma,
 			return -EINVAL;
 		}
 		flags = new_flags;
+		vma->vm_page_prot = __pgprot(
+			(pgprot_val(vma->vm_page_prot) & ~_PAGE_CACHE_MASK) |
+			flags);
 	}
 
 	if (((vma->vm_pgoff < max_low_pfn_mapped) ||
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 065cdf8..3daa05f 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -98,7 +98,7 @@ extern unsigned int kobjsize(const void *objp);
 #define VM_HUGETLB	0x00400000	/* Huge TLB Page VM */
 #define VM_NONLINEAR	0x00800000	/* Is non-linear (remap_file_pages) */
 #define VM_MAPPED_COPY	0x01000000	/* T if mapped copy of data (nommu mmap) */
-#define VM_INSERTPAGE	0x02000000	/* The vma has had "vm_insert_page()" done on it */
+#define VM_INSERTPAGE	0x02000000	/* The vma has had "vm_insert_page()" done on it. Refer note in VM_PFNMAP_AT_MMAP below */
 #define VM_ALWAYSDUMP	0x04000000	/* Always include in core dumps */
 
 #define VM_CAN_NONLINEAR 0x08000000	/* Has ->fault & does nonlinear pages */
@@ -127,6 +127,17 @@ extern unsigned int kobjsize(const void *objp);
 #define VM_SPECIAL (VM_IO | VM_DONTEXPAND | VM_RESERVED | VM_PFNMAP)
 
 /*
+ * pfnmap vmas that are fully mapped at mmap time (not mapped on fault).
+ * Used by x86 PAT to identify such PFNMAP mappings and optimize their handling.
+ * Note VM_INSERTPAGE flag is overloaded here. i.e,
+ * VM_INSERTPAGE && !VM_PFNMAP implies
+ *     The vma has had "vm_insert_page()" done on it
+ * VM_INSERTPAGE && VM_PFNMAP implies
+ *     The vma is PFNMAP with full mapping at mmap time
+ */
+#define VM_PFNMAP_AT_MMAP (VM_INSERTPAGE | VM_PFNMAP)
+
+/*
  * mapping from the currently active vm_flags protection bits (the
  * low four bits) to a page protection mask..
  */
@@ -145,7 +156,7 @@ extern pgprot_t protection_map[16];
  */
 static inline int is_linear_pfn_mapping(struct vm_area_struct *vma)
 {
-	return ((vma->vm_flags & VM_PFNMAP) && vma->vm_pgoff);
+	return ((vma->vm_flags & VM_PFNMAP_AT_MMAP) == VM_PFNMAP_AT_MMAP);
 }
 
 static inline int is_pfn_mapping(struct vm_area_struct *vma)
diff --git a/mm/memory.c b/mm/memory.c
index baa999e..d7df5ba 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -1665,9 +1665,10 @@ int remap_pfn_range(struct vm_area_struct *vma, unsigned long addr,
 	 * behaviour that some programs depend on. We mark the "original"
 	 * un-COW'ed pages by matching them up with "vma->vm_pgoff".
 	 */
-	if (addr == vma->vm_start && end == vma->vm_end)
+	if (addr == vma->vm_start && end == vma->vm_end) {
 		vma->vm_pgoff = pfn;
-	else if (is_cow_mapping(vma->vm_flags))
+		vma->vm_flags |= VM_PFNMAP_AT_MMAP;
+	} else if (is_cow_mapping(vma->vm_flags))
 		return -EINVAL;
 
 	vma->vm_flags |= VM_IO | VM_RESERVED | VM_PFNMAP;
@@ -1679,6 +1680,7 @@ int remap_pfn_range(struct vm_area_struct *vma, unsigned long addr,
 		 * needed from higher level routine calling unmap_vmas
 		 */
 		vma->vm_flags &= ~(VM_IO | VM_RESERVED | VM_PFNMAP);
+		vma->vm_flags &= ~VM_PFNMAP_AT_MMAP;
 		return -EINVAL;
 	}
 
-- 
1.6.2



Index: TODO
===================================================================
RCS file: /cvs/pkgs/rpms/kernel/devel/TODO,v
retrieving revision 1.60
retrieving revision 1.61
diff -u -r1.60 -r1.61
--- TODO	30 Mar 2009 08:56:09 -0000	1.60
+++ TODO	30 Mar 2009 15:08:54 -0000	1.61
@@ -96,3 +96,7 @@
         virtio_net guest->remote GSO busted with 2.6.29 host
         https://bugzilla.redhat.com/490266
         Should be in 2.6.29.1
+
+* linux-2.6.29-pat-fixes.patch:
+	http://bugs.freedesktop.org/show_bug.cgi?id=20803
+	ajax to follow up with jbarnes


Index: kernel.spec
===================================================================
RCS file: /cvs/pkgs/rpms/kernel/devel/kernel.spec,v
retrieving revision 1.1484
retrieving revision 1.1485
diff -u -r1.1484 -r1.1485
--- kernel.spec	30 Mar 2009 12:11:43 -0000	1.1484
+++ kernel.spec	30 Mar 2009 15:08:54 -0000	1.1485
@@ -701,6 +701,9 @@
 # fix locking in ipsec (#489764)
 Patch9200: linux-2.6-net-xfrm-fix-spin-unlock.patch
 
+# http://bugs.freedesktop.org/show_bug.cgi?id=20803
+Patch9210: linux-2.6.29-pat-fixes.patch
+
 %endif
 
 BuildRoot: %{_tmppath}/kernel-%{KVERREL}-root
@@ -1261,6 +1264,8 @@
 
 ApplyPatch linux-2.6-net-xfrm-fix-spin-unlock.patch
 
+ApplyPatch linux-2.6.29-pat-fixes.patch
+
 # END OF PATCH APPLICATIONS
 
 %endif
@@ -1844,6 +1849,9 @@
 # and build.
 
 %changelog
+* Mon Mar 30 2009 Adam Jackson <ajax at redhat.com>
+- linux-2.6.29-pat-fixes.patch: Fix PAT/GTT interaction
+
 * Mon Mar 30 2009 Mauro Carvalho Chehab <mchehab at redhat.com>
 - some fixes of troubles caused by v4l2 subdev conversion