]> www.infradead.org Git - users/jedix/linux-maple.git/commit
x86/mm: Make e820__end_ram_pfn() cover E820_TYPE_ACPI ranges
authorKirill A. Shutemov <kirill.shutemov@linux.intel.com>
Fri, 14 Jun 2024 09:58:57 +0000 (12:58 +0300)
committerBorislav Petkov (AMD) <bp@alien8.de>
Mon, 17 Jun 2024 15:46:08 +0000 (17:46 +0200)
commit06fa48d85b09b3e67afeda220bc19f7102b53beb
tree4bf5652af251c807c0193f1176f02f96eae5a5fe
parent859e63b789d6b17b3c64e51a0aabdc58752a0254
x86/mm: Make e820__end_ram_pfn() cover E820_TYPE_ACPI ranges

e820__end_of_ram_pfn() is used to calculate max_pfn which, among other things,
guides where direct mapping ends. Any memory above max_pfn is not going to be
present in the direct mapping.

e820__end_of_ram_pfn() finds the end of the RAM based on the highest
E820_TYPE_RAM range. But it doesn't includes E820_TYPE_ACPI ranges into
calculation.

Despite the name, E820_TYPE_ACPI covers not only ACPI data, but also EFI tables
and might be required by kernel to function properly.

Usually the problem is hidden because there is some E820_TYPE_RAM memory above
E820_TYPE_ACPI. But crashkernel only presents pre-allocated crash memory as
E820_TYPE_RAM on boot. If the pre-allocated range is small, it can fit under the
last E820_TYPE_ACPI range.

Modify e820__end_of_ram_pfn() and e820__end_of_low_ram_pfn() to cover
E820_TYPE_ACPI memory.

The problem was discovered during debugging kexec for TDX guest. TDX guest uses
E820_TYPE_ACPI to store the unaccepted memory bitmap and pass it between the
kernels on kexec.

Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Dave Hansen <dave.hansen@intel.com>
Tested-by: Tao Liu <ltao@redhat.com>
Link: https://lore.kernel.org/r/20240614095904.1345461-13-kirill.shutemov@linux.intel.com
arch/x86/kernel/e820.c