linux-kernel - [PATCH 3.16 207/346] powerpc/prom: Fix sub-processor option passed to ibm, client-architecture-support

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Mon, 14 Nov 2016 00:14:20 +0000
From:   Ben Hutchings <ben@...adent.org.uk>
To:     linux-kernel@...r.kernel.org, stable@...r.kernel.org
CC:     akpm@...ux-foundation.org,
        "Benjamin Herrenschmidt" <benh@...nel.crashing.org>,
        "Michael Ellerman" <mpe@...erman.id.au>
Subject: [PATCH 3.16 207/346] powerpc/prom: Fix sub-processor option
 passed to ibm, client-architecture-support

3.16.39-rc1 review patch.  If anyone has any objections, please let me know.

------------------

From: Michael Ellerman <mpe@...erman.id.au>

commit 66443efa83dc73775100b7442962ce2cb0d4472e upstream.

When booting from an OpenFirmware which supports it, we use the
"ibm,client-architecture-support" firmware call to communicate
our capabilities to firmware.

The format of the structure we pass to firmware is specified in
PAPR (Power Architecture Platform Requirements), or the public version
LoPAPR (Linux on Power Architecture Platform Reference).

Referring to table 244 in LoPAPR v1.1, option vector 5 contains a 4 byte
field at bytes 17-20 for the "Platform Facilities Enable". This is
followed by a 1 byte field at byte 21 for "Sub-Processor Represenation
Level".

Comparing to the code, there we have the Platform Facilities
options (OV5_PFO_*) at byte 17, but we fail to pad that field out to its
full width of 4 bytes. This means the OV5_SUB_PROCESSORS option is
incorrectly placed at byte 18.

Fix it by adding zero bytes for bytes 18, 19, 20, and comment the bytes
to hopefully make it clearer in future.

As far as I'm aware nothing actually consumes this value at this time,
so the effect of this bug is nil in practice.

It does mean we've been incorrectly setting bit 15 of the "Platform
Facilities Enable" option for the past ~3 1/2 years, so we should avoid
allocating that bit to anything else in future.

Fixes: df77c7992029 ("powerpc/pseries: Update ibm,architecture.vec for PAPR 2.7/POWER8")
Signed-off-by: Michael Ellerman <mpe@...erman.id.au>
Signed-off-by: Benjamin Herrenschmidt <benh@...nel.crashing.org>
[bwh: Backported to 3.16:
 - Adjust context
 - Length calculations don't use VECTOR_LENGTH()]
Signed-off-by: Ben Hutchings <ben@...adent.org.uk>
---
 arch/powerpc/kernel/prom_init.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

--- a/arch/powerpc/kernel/prom_init.c
+++ b/arch/powerpc/kernel/prom_init.c
@@ -685,7 +685,7 @@ unsigned char ibm_architecture_vec[] = {
 	OV4_MIN_ENT_CAP,		/* minimum VP entitled capacity */

 	/* option vector 5: PAPR/OF options */
-	19 - 2,				/* length */
+	22 - 2,				/* length */
 	0,				/* don't ignore, don't halt */
 	OV5_FEAT(OV5_LPAR) | OV5_FEAT(OV5_SPLPAR) | OV5_FEAT(OV5_LARGE_PAGES) |
 	OV5_FEAT(OV5_DRCONF_MEMORY) | OV5_FEAT(OV5_DONATE_DEDICATE_CPU) |
@@ -716,8 +716,11 @@ unsigned char ibm_architecture_vec[] = {
 	0,
 	0,
 	OV5_FEAT(OV5_PFO_HW_RNG) | OV5_FEAT(OV5_PFO_HW_ENCR) |
-	OV5_FEAT(OV5_PFO_HW_842),
-	OV5_FEAT(OV5_SUB_PROCESSORS),
+	OV5_FEAT(OV5_PFO_HW_842),				/* Byte 17 */
+	0,							/* Byte 18 */
+	0,							/* Byte 19 */
+	0,							/* Byte 20 */
+	OV5_FEAT(OV5_SUB_PROCESSORS),				/* Byte 21 */
 	/* option vector 6: IBM PAPR hints */
 	4 - 2,				/* length */
 	0,