hw/ssi/xilinx_spips: Avoid AXI writes to the LQSPI linear memory
PATCH v2 1/1 hw/ssi/xilinx_spips: Avoid AXI writes to the LQSPI linear memory
Lei Sun found while auditing the code than a CPU write would trigger a NULL pointer deference.
From UG1085 datasheet [*] AXI writes in this region are ignored and generates an External Slave Error (SLVERR).
Fix by checking the access is a READ before calling the region callback.
[*] https://www.xilinx.com/support/documentation/user_guides/ug1085-zynq-ultrascale-trm.pdf
---
v1: https://lists.gnu.org/archive/html/qemu-devel/2019-07/msg01238.html
--- hw/ssi/xilinx_spips.c | 13 +++++++++++++ 1 file changed, 13 insertions(+)
hw/ssi/xilinx_spips.c +13 -0
@@ -1202,6 +1202,18 @@ static void lqspi_load_cache(void *opaque, hwaddr addr) } } +static bool lqspi_accepts (void *opaque,hwaddr addr, +unsigned size,bool is_write, +MemTxAttrs attrs) +{ +/* +* From UG1085, Chapter 24 (Quad-SPI controllers): +* - Writes are ignored +* - AXI writes generate an external AXI slave error (SLVERR) +*/ +return !is_write; +} +static uint64_t lqspi_read (void *opaque,hwaddr addr,unsigned int size) {
@@ -1225,6 +1237,7 @@ static const MemoryRegionOps lqspi_ops = { .read = lqspi_read, .endianness = DEVICE_NATIVE_ENDIAN, .valid = { + .accepts = lqspi_accepts, .min_access_size = 1, .max_access_size =4 }
Signature · 1 line
2.20.1
+-- On Fri, 5 Jul 2019, Philippe Mathieu-Daudé wrote --+
| +static bool lqspi_accepts(void *opaque, hwaddr addr,
| + unsigned size, bool is_write,
| + MemTxAttrs attrs)
| +{
| + /*
| + * From UG1085, Chapter 24 (Quad-SPI controllers):
| + * - Writes are ignored
| + * - AXI writes generate an external AXI slave error (SLVERR)
| + */
| + return !is_write;
| +}
| +
| static uint64_t
| lqspi_read(void *opaque, hwaddr addr, unsigned int size)
| {
| @@ -1225,6 +1237,7 @@ static const MemoryRegionOps lqspi_ops = {
| .read = lqspi_read,
| .endianness = DEVICE_NATIVE_ENDIAN,
| .valid = {
| + .accepts = lqspi_accepts,
| .min_access_size = 1,
| .max_access_size = 4
| }Looks okay.
To confirm,
When lqspi_accepts() returns false, guest will see an error/exception?
Thank you.
--
Prasad J Pandit / Red Hat Product Security Team
47AF CE69 3A90 54AA 9045 1053 DD13 3D32 FE5B 041F
Hi Philippe,
Quoted 27 lines · Philippe Mathieu-Daudé
Lei Sun found while auditing the code than a CPU write would trigger a NULL pointer deference.
From UG1085 datasheet [*] AXI writes in this region are ignored and generates an External Slave Error (SLVERR).
Fix by checking the access is a READ before calling the region callback.
[*] https://www.xilinx.com/support/documentation/user_guides/ug1085-zynq-ultrascale-trm.pdf
Reported-by Lei Sun <[email protected]>Signed-off-by Philippe Mathieu-Daudé <[email protected]>---
v1: https://lists.gnu.org/archive/html/qemu-devel/2019-07/msg01238.html--- hw/ssi/xilinx_spips.c | 13 +++++++++++++ 1 file changed, 13 insertions(+)hw/ssi/xilinx_spips.c +0
-0@@ -1202,6 +1202,18 @@ static void lqspi_load_cache(void *opaque, hwaddr addr) } }
hw/ssi/xilinx_spips.c static void lqspi_load_cache(void *opaque, hwaddr addr) · line 1214
Quoted 10 lines
hw/ssi/xilinx_spips.c +10
-0+static bool lqspi_accepts (void *opaque,hwaddr addr, +unsigned size,bool is_write, +MemTxAttrs attrs) +{ +/* +* From UG1085, Chapter 24 (Quad-SPI controllers): +* - Writes are ignored +* - AXI writes generate an external AXI slave error (SLVERR) +*/ +return !is_write;
What I can see above generates an AXI Decode error (MEMTX_DECODE_ERROR), MEMTX_ERROR is the one that maps to AXI slave error. What about swapping to the solution you proposed yesterday? (but changing return value of the write function to be MEMTX_ERROR)
Best regards,
Francisco Iglesias
Quoted 16 lines
hw/ssi/xilinx_spips.c +3
-0+} +static uint64_t lqspi_read (void *opaque,hwaddr addr,unsigned int size) {@@ -1225,6 +1237,7 @@ static const MemoryRegionOps lqspi_ops = { .read = lqspi_read, .endianness = DEVICE_NATIVE_ENDIAN, .valid = { + .accepts = lqspi_accepts, .min_access_size = 1, .max_access_size =4 }Signature · 1 line
2.20.1