pstore/ram: Correctly calculate usable PRZ bytes
authorKees Cook <keescook@chromium.org>
Thu, 1 Nov 2018 23:17:22 +0000 (16:17 -0700)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Mon, 17 Dec 2018 08:28:53 +0000 (09:28 +0100)
commitb718d6bede4bd6c8c6faac1369567f3d7587d4c1
tree83f4471e7433b959fd12cf4fe642de1a2cc4c7c5
parentb9c242b4fad864505f679db31531903d1b3779e9
pstore/ram: Correctly calculate usable PRZ bytes

[ Upstream commit 89d328f637b9904b6d4c9af73c8a608b8dd4d6f8 ]

The actual number of bytes stored in a PRZ is smaller than the
bytes requested by platform data, since there is a header on each
PRZ. Additionally, if ECC is enabled, there are trailing bytes used
as well. Normally this mismatch doesn't matter since PRZs are circular
buffers and the leading "overflow" bytes are just thrown away. However, in
the case of a compressed record, this rather badly corrupts the results.

This corruption was visible with "ramoops.mem_size=204800 ramoops.ecc=1".
Any stored crashes would not be uncompressable (producing a pstorefs
"dmesg-*.enc.z" file), and triggering errors at boot:

  [    2.790759] pstore: crypto_comp_decompress failed, ret = -22!

Backporting this depends on commit 70ad35db3321 ("pstore: Convert console
write to use ->write_buf")

Reported-by: Joel Fernandes <joel@joelfernandes.org>
Fixes: b0aad7a99c1d ("pstore: Add compression support to pstore")
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
fs/pstore/ram.c
include/linux/pstore.h