]> git.ipfire.org Git - thirdparty/kernel/stable.git/commit
riscv: Properly export reserved regions in /proc/iomem
authorBjörn Töpel <bjorn@rivosinc.com>
Wed, 9 Apr 2025 18:21:27 +0000 (20:21 +0200)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Fri, 25 Apr 2025 08:45:45 +0000 (10:45 +0200)
commitd7c65ecad9596c0e9c58cc26ee805031ba00df4d
treed8c86b8500c0100a6fc9c7d51cfe2892c828954f
parenta2874f0dff63829d1f540003e2d83adb610ee64a
riscv: Properly export reserved regions in /proc/iomem

[ Upstream commit e94eb7ea6f206e229791761a5fdf9389f8dbd183 ]

The /proc/iomem represents the kernel's memory map. Regions marked
with "Reserved" tells the user that the range should not be tampered
with. Kexec-tools, when using the older kexec_load syscall relies on
the "Reserved" regions to build the memory segments, that will be the
target of the new kexec'd kernel.

The RISC-V port tries to expose all reserved regions to userland, but
some regions were not properly exposed: Regions that resided in both
the "regular" and reserved memory block, e.g. the EFI Memory Map. A
missing entry could result in reserved memory being overwritten.

It turns out, that arm64, and loongarch had a similar issue a while
back:

  commit d91680e687f4 ("arm64: Fix /proc/iomem for reserved but not memory regions")
  commit 50d7ba36b916 ("arm64: export memblock_reserve()d regions via /proc/iomem")

Similar to the other ports, resolve the issue by splitting the regions
in an arch initcall, since we need a working allocator.

Fixes: ffe0e5261268 ("RISC-V: Improve init_resources()")
Signed-off-by: Björn Töpel <bjorn@rivosinc.com>
Reviewed-by: Alexandre Ghiti <alexghiti@rivosinc.com>
Link: https://lore.kernel.org/r/20250409182129.634415-1-bjorn@kernel.org
Signed-off-by: Alexandre Ghiti <alexghiti@rivosinc.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
arch/riscv/kernel/setup.c