x86/ioremap: Fix page aligned size calculation in __ioremap_caller()
authorMichael Kelley <mikelley@microsoft.com>
Wed, 16 Nov 2022 18:41:24 +0000 (10:41 -0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Fri, 2 Dec 2022 16:40:04 +0000 (17:40 +0100)
commit22870431cd250df3eec96753d9422bfc9f0d52a4
tree5662eb8d42dcfe0dfc99a5a846a84ccbe9899461
parent3fdeacf087ff92c85162b8a0e111dfe2479238ac
x86/ioremap: Fix page aligned size calculation in __ioremap_caller()

commit 4dbd6a3e90e03130973688fd79e19425f720d999 upstream.

Current code re-calculates the size after aligning the starting and
ending physical addresses on a page boundary. But the re-calculation
also embeds the masking of high order bits that exceed the size of
the physical address space (via PHYSICAL_PAGE_MASK). If the masking
removes any high order bits, the size calculation results in a huge
value that is likely to immediately fail.

Fix this by re-calculating the page-aligned size first. Then mask any
high order bits using PHYSICAL_PAGE_MASK.

Fixes: ffa71f33a820 ("x86, ioremap: Fix incorrect physical address handling in PAE mode")
Signed-off-by: Michael Kelley <mikelley@microsoft.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Dave Hansen <dave.hansen@linux.intel.com>
Cc: <stable@kernel.org>
Link: https://lore.kernel.org/r/1668624097-14884-2-git-send-email-mikelley@microsoft.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
arch/x86/mm/ioremap.c