]> git.baikalelectronics.ru Git - kernel.git/commit
arm[64]/memremap: don't abuse pfn_valid() to ensure presence of linear map
authorMike Rapoport <rppt@linux.ibm.com>
Tue, 10 May 2022 00:34:28 +0000 (17:34 -0700)
committerakpm <akpm@linux-foundation.org>
Tue, 10 May 2022 00:34:28 +0000 (17:34 -0700)
commit0b69b51c8e03287d13525fc212863ab1992e4482
tree7e078fa82423ad47b7b6d8932633b562c3377ed9
parentc2a652ab17eb50de5d9252fa486d85430c81b2b5
arm[64]/memremap: don't abuse pfn_valid() to ensure presence of linear map

The semantics of pfn_valid() is to check presence of the memory map for a
PFN and not whether a PFN is covered by the linear map.  The memory map
may be present for NOMAP memory regions, but they won't be mapped in the
linear mapping.  Accessing such regions via __va() when they are
memremap()'ed will cause a crash.

On v5.4.y the crash happens on qemu-arm with UEFI [1]:

<1>[    0.084476] 8<--- cut here ---
<1>[    0.084595] Unable to handle kernel paging request at virtual address dfb76000
<1>[    0.084938] pgd = (ptrval)
<1>[    0.085038] [dfb76000] *pgd=5f7fe801, *pte=00000000, *ppte=00000000

...

<4>[    0.093923] [<c0ed6ce8>] (memcpy) from [<c16a06f8>] (dmi_setup+0x60/0x418)
<4>[    0.094204] [<c16a06f8>] (dmi_setup) from [<c16a38d4>] (arm_dmi_init+0x8/0x10)
<4>[    0.094408] [<c16a38d4>] (arm_dmi_init) from [<c0302e9c>] (do_one_initcall+0x50/0x228)
<4>[    0.094619] [<c0302e9c>] (do_one_initcall) from [<c16011e4>] (kernel_init_freeable+0x15c/0x1f8)
<4>[    0.094841] [<c16011e4>] (kernel_init_freeable) from [<c0f028cc>] (kernel_init+0x8/0x10c)
<4>[    0.095057] [<c0f028cc>] (kernel_init) from [<c03010e8>] (ret_from_fork+0x14/0x2c)

On kernels v5.10.y and newer the same crash won't reproduce on ARM because
commit 60c9bc1b5f05 ("arch, drivers: replace for_each_membock() with
for_each_mem_range()") changed the way memory regions are registered in
the resource tree, but that merely covers up the problem.

On ARM64 memory resources registered in yet another way and there the
issue of wrong usage of pfn_valid() to ensure availability of the linear
map is also covered.

Implement arch_memremap_can_ram_remap() on ARM and ARM64 to prevent access
to NOMAP regions via the linear mapping in memremap().

Link: https://lore.kernel.org/all/Yl65zxGgFzF1Okac@sirena.org.uk
Link: https://lkml.kernel.org/r/20220426060107.7618-1-rppt@kernel.org
Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Reported-by: "kernelci.org bot" <bot@kernelci.org>
Tested-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Mark Brown <broonie@kernel.org>
Cc: Mark-PK Tsai <mark-pk.tsai@mediatek.com>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Tony Lindgren <tony@atomide.com>
Cc: Will Deacon <will@kernel.org>
Cc: <stable@vger.kernel.org> [5.4+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
arch/arm/include/asm/io.h
arch/arm/mm/ioremap.c
arch/arm64/include/asm/io.h
arch/arm64/mm/ioremap.c