x86/pci: Check signature before assigning shadow ROM
PATCH v2 1/1 x86/pci: Check signature before assigning shadow ROM
Recent IGD platforms without VBIOS or UEFI CSM support do not contain VGA ROM at 0xC0000. Check whether the VGA ROM region is a valid PCI option ROM with 0xAA55 signature before assigning the shadow ROM to the default PCI VGA controller.
---
Changelog:
v2:
* Use memmap() instead of iomap() as the shadow ROM is copied to RAM by BIOS
* Only map the first 2 bytes for the signature check.
arch/x86/pci/fixup.c | 12 ++++++++++++ 1 file changed, 12 insertions(+)
arch/x86/pci/fixup.c +12 -0
@@ -357,6 +357,18 @@ static void pci_fixup_video(struct pci_dev *pdev)struct pci_bus *bus;u16 config;struct resource *res; +void *rom; +u16 sig; + +/* Does VBIOS region contain a valid PCI ROM? */ + rom =memremap (0xC0000 ,sizeof (sig),MEMREMAP_WB ); +if (!rom) +return ; + +memcpy (&sig, rom,sizeof (sig)); +memunmap (rom); +if (sig !=0xAA55 ) +return ;/* Is VGA routed to us? */ bus = pdev->bus;
Signature · 1 line
2.51.0
Quoted 4 lines · Tomita Moeko
Recent IGD platforms without VBIOS or UEFI CSM support do not contain VGA ROM at
0xC0000. Check whether the VGA ROM region is a valid PCI option ROM with0xAA55signature before assigning the shadow ROM to the default PCI VGA controller.
[...]
Quoted 10 lines
+++
b/arch/x86/pci/fixup.carch/x86/pci/fixup.c +5
-0@@ -357,6 +357,18 @@ static void pci_fixup_video(struct pci_dev *pdev)struct pci_bus *bus;u16 config;struct resource *res; +void *rom; +u16 sig; + +/* Does VBIOS region contain a valid PCI ROM? */ + rom =memremap (0xC0000 ,sizeof (sig),MEMREMAP_WB );
arch/x86/pci/fixup.c static void pci_fixup_video(struct pci_dev *pdev) · line 374
Quoted 10 lines
arch/x86/pci/fixup.c +7
-0+if (!rom) +return ; + +memcpy (&sig, rom,sizeof (sig)); +memunmap (rom); +if (sig !=0xAA55 ) +return ;/* Is VGA routed to us? */ bus = pdev->bus;
I have to ask again, in arch/x86/kernel/probe_roms.c:probe_roms(), the signature is already verified. If it doesn't match, the video_rom_resource isn't added to iomem_resource.
Which makes me wonder, wouldn't it be sufficient to just do something like:
if (!lookup_resource(&iomem_resource, 0xC0000)) return;
Another thought I have, I'd move the code you're inserting further down, perhaps after the while-loop. Actually the existing code isn't very pretty, there should be a return after failure of the vga_default_device checks and after the Command register check so that the actual resource adjustment doesn't need to be indented.
Thanks,
Lukas