Re: vm_area at addr ffffffffc0800000 is not marked as VM_IOREMAP

[Alexei Starovoitov <alexei.starovoitov@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 695 bytes --]

On Fri, Mar 8, 2024 at 9:24 AM Christoph Hellwig <hch@infradead.org> wrote:
>
> On Fri, Mar 08, 2024 at 09:20:24AM -0800, Alexei Starovoitov wrote:
> > ok. Like the attached patch?
>
> Looks sensibe, but I think the powerpc callers of ioremap_page_range
> will need the same treatment.

Good point. Only one of the callers in arch/powerpc needs adjusting.
Found few other similar arch users.
See attached patch.

ioremap_page() in arch/arm/mm/ioremap.c is an interesting case.
It is EXPORT_SYMBOL, but there are no in-tree users.
I think we shouldn't apply checks to it,
since some out-of-tree module may fail.
I have no arm boards to test, I suggest we play safe than sorry.

[-- Attachment #2: 0001-mm-Introduce-vmap_page_range-to-map-pages-in-PCI-add.patch --]
[-- Type: application/octet-stream, Size: 6504 bytes --]

From bf82806a2800b49c5627bbf057406245d0717ccd Mon Sep 17 00:00:00 2001
From: Alexei Starovoitov <ast@kernel.org>
Date: Fri, 8 Mar 2024 09:12:54 -0800
Subject: [PATCH bpf-next] mm: Introduce vmap_page_range() to map pages in PCI
 address space

ioremap_page_range() should be used for ranges within vmalloc range only.
The vmalloc ranges are allocated by get_vm_area(). PCI has "resource"
allocator that manages PCI_IOBASE, IO_SPACE_LIMIT address range, hence
introduce vmap_page_range() to be used exclusively to map pages
in PCI address space.

Fixes: 3e49a866c9dc ("mm: Enforce VM_IOREMAP flag and range in ioremap_page_range.")
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
---
 arch/arm/mm/ioremap.c            |  8 ++++----
 arch/loongarch/kernel/setup.c    |  2 +-
 arch/mips/loongson64/init.c      |  2 +-
 arch/powerpc/kernel/isa-bridge.c |  4 ++--
 drivers/pci/pci.c                |  4 ++--
 include/linux/io.h               |  7 +++++++
 mm/vmalloc.c                     | 23 +++++++++++++++--------
 7 files changed, 32 insertions(+), 18 deletions(-)

diff --git a/arch/arm/mm/ioremap.c b/arch/arm/mm/ioremap.c
index 2129070065c3..794cfea9f9d4 100644
--- a/arch/arm/mm/ioremap.c
+++ b/arch/arm/mm/ioremap.c
@@ -110,8 +110,8 @@ void __init add_static_vm_early(struct static_vm *svm)
 int ioremap_page(unsigned long virt, unsigned long phys,
 		 const struct mem_type *mtype)
 {
-	return ioremap_page_range(virt, virt + PAGE_SIZE, phys,
-				  __pgprot(mtype->prot_pte));
+	return vmap_page_range(virt, virt + PAGE_SIZE, phys,
+			       __pgprot(mtype->prot_pte));
 }
 EXPORT_SYMBOL(ioremap_page);
 
@@ -466,8 +466,8 @@ int pci_remap_iospace(const struct resource *res, phys_addr_t phys_addr)
 	if (res->end > IO_SPACE_LIMIT)
 		return -EINVAL;
 
-	return ioremap_page_range(vaddr, vaddr + resource_size(res), phys_addr,
-				  __pgprot(get_mem_type(pci_ioremap_mem_type)->prot_pte));
+	return vmap_page_range(vaddr, vaddr + resource_size(res), phys_addr,
+			       __pgprot(get_mem_type(pci_ioremap_mem_type)->prot_pte));
 }
 EXPORT_SYMBOL(pci_remap_iospace);
 
diff --git a/arch/loongarch/kernel/setup.c b/arch/loongarch/kernel/setup.c
index 634ef17fd38b..fd915ad69c09 100644
--- a/arch/loongarch/kernel/setup.c
+++ b/arch/loongarch/kernel/setup.c
@@ -490,7 +490,7 @@ static int __init add_legacy_isa_io(struct fwnode_handle *fwnode,
 	}
 
 	vaddr = (unsigned long)(PCI_IOBASE + range->io_start);
-	ioremap_page_range(vaddr, vaddr + size, hw_start, pgprot_device(PAGE_KERNEL));
+	vmap_page_range(vaddr, vaddr + size, hw_start, pgprot_device(PAGE_KERNEL));
 
 	return 0;
 }
diff --git a/arch/mips/loongson64/init.c b/arch/mips/loongson64/init.c
index 553142c1f14f..a35dd7311795 100644
--- a/arch/mips/loongson64/init.c
+++ b/arch/mips/loongson64/init.c
@@ -180,7 +180,7 @@ static int __init add_legacy_isa_io(struct fwnode_handle *fwnode, resource_size_
 
 	vaddr = PCI_IOBASE + range->io_start;
 
-	ioremap_page_range(vaddr, vaddr + size, hw_start, pgprot_device(PAGE_KERNEL));
+	vmap_page_range(vaddr, vaddr + size, hw_start, pgprot_device(PAGE_KERNEL));
 
 	return 0;
 }
diff --git a/arch/powerpc/kernel/isa-bridge.c b/arch/powerpc/kernel/isa-bridge.c
index 48e0eaf1ad61..5c064485197a 100644
--- a/arch/powerpc/kernel/isa-bridge.c
+++ b/arch/powerpc/kernel/isa-bridge.c
@@ -46,8 +46,8 @@ static void remap_isa_base(phys_addr_t pa, unsigned long size)
 	WARN_ON_ONCE(size & ~PAGE_MASK);
 
 	if (slab_is_available()) {
-		if (ioremap_page_range(ISA_IO_BASE, ISA_IO_BASE + size, pa,
-				pgprot_noncached(PAGE_KERNEL)))
+		if (vmap_page_range(ISA_IO_BASE, ISA_IO_BASE + size, pa,
+				    pgprot_noncached(PAGE_KERNEL)))
 			vunmap_range(ISA_IO_BASE, ISA_IO_BASE + size);
 	} else {
 		early_ioremap_range(ISA_IO_BASE, pa, size,
diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index c3585229c12a..ccee56615f78 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -4353,8 +4353,8 @@ int pci_remap_iospace(const struct resource *res, phys_addr_t phys_addr)
 	if (res->end > IO_SPACE_LIMIT)
 		return -EINVAL;
 
-	return ioremap_page_range(vaddr, vaddr + resource_size(res), phys_addr,
-				  pgprot_device(PAGE_KERNEL));
+	return vmap_page_range(vaddr, vaddr + resource_size(res), phys_addr,
+			       pgprot_device(PAGE_KERNEL));
 #else
 	/*
 	 * This architecture does not have memory mapped I/O space,
diff --git a/include/linux/io.h b/include/linux/io.h
index 7304f2a69960..235ba7d80a8f 100644
--- a/include/linux/io.h
+++ b/include/linux/io.h
@@ -23,12 +23,19 @@ void __iowrite64_copy(void __iomem *to, const void *from, size_t count);
 #ifdef CONFIG_MMU
 int ioremap_page_range(unsigned long addr, unsigned long end,
 		       phys_addr_t phys_addr, pgprot_t prot);
+int vmap_page_range(unsigned long addr, unsigned long end,
+		    phys_addr_t phys_addr, pgprot_t prot);
 #else
 static inline int ioremap_page_range(unsigned long addr, unsigned long end,
 				     phys_addr_t phys_addr, pgprot_t prot)
 {
 	return 0;
 }
+static inline int vmap_page_range(unsigned long addr, unsigned long end,
+				  phys_addr_t phys_addr, pgprot_t prot)
+{
+	return 0;
+}
 #endif
 
 /*
diff --git a/mm/vmalloc.c b/mm/vmalloc.c
index e5b8c70950bc..1e36322d83d8 100644
--- a/mm/vmalloc.c
+++ b/mm/vmalloc.c
@@ -304,11 +304,24 @@ static int vmap_range_noflush(unsigned long addr, unsigned long end,
 	return err;
 }
 
+int vmap_page_range(unsigned long addr, unsigned long end,
+		    phys_addr_t phys_addr, pgprot_t prot)
+{
+	int err;
+
+	err = vmap_range_noflush(addr, end, phys_addr, pgprot_nx(prot),
+				 ioremap_max_page_shift);
+	flush_cache_vmap(addr, end);
+	if (!err)
+		err = kmsan_ioremap_page_range(addr, end, phys_addr, prot,
+					       ioremap_max_page_shift);
+	return err;
+}
+
 int ioremap_page_range(unsigned long addr, unsigned long end,
 		phys_addr_t phys_addr, pgprot_t prot)
 {
 	struct vm_struct *area;
-	int err;
 
 	area = find_vm_area((void *)addr);
 	if (!area || !(area->flags & VM_IOREMAP)) {
@@ -322,13 +335,7 @@ int ioremap_page_range(unsigned long addr, unsigned long end,
 			  (long)area->addr + get_vm_area_size(area));
 		return -ERANGE;
 	}
-	err = vmap_range_noflush(addr, end, phys_addr, pgprot_nx(prot),
-				 ioremap_max_page_shift);
-	flush_cache_vmap(addr, end);
-	if (!err)
-		err = kmsan_ioremap_page_range(addr, end, phys_addr, prot,
-					       ioremap_max_page_shift);
-	return err;
+	return vmap_page_range(addr, end, phys_addr, prot);
 }
 
 static void vunmap_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end,
-- 
2.43.0