Skip to contentSkip to conversation

dma: bestcomm: gen_bd: split struct bcom_psc_params from array definition

PATCH Superseded Rosen Penev 2 participants 3 messages May 30 dmaengine

lore
Share & tools
lore
#0 lore raw

PATCH v2 1/1 dma: bestcomm: gen_bd: split struct bcom_psc_params from array definition

The combined struct-definition-with-initializer pattern confuses the kernel-doc parser. Split into separate struct definition and array declaration.

Also now that it's fixed, it warns on missing members. Add those as well.

Since this is just a lookup table and not modified, make it const so that it can be moved to read only memory.

Assisted-by Opencode:Big-pickle

---
v2: add const and add variable descriptions

 drivers/dma/bestcomm/gen_bd.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)
drivers/dma/bestcomm/gen_bd.c +9 -3
index 8a24a5cbc263..61b5746e1a97 100644
@@ -254,17 +254,23 @@ EXPORT_SYMBOL_GPL(bcom_gen_bd_tx_release);
  */
 
 /**
- * bcom_psc_parameters - Bestcomm initialization value table for PSC devices
+ * struct bcom_psc_params - Bestcomm initialization value table for PSC devices
+ * @rx_initiator: RX initiator ID
+ * @rx_ipr: RX interrupt priority register value
+ * @tx_initiator: TX initiator ID
+ * @tx_ipr: TX interrupt priority register value
  *
  * This structure is only used internally.  It is a lookup table for PSC
  * specific parameters to bestcomm tasks.
  */
-static struct bcom_psc_params {
+struct bcom_psc_params {
 	int rx_initiator;
 	int rx_ipr;
 	int tx_initiator;
 	int tx_ipr;
-} bcom_psc_params[] = {
+};
+
+static const struct bcom_psc_params bcom_psc_params[] = {
 	[0] = {
 		.rx_initiator = BCOM_INITIATOR_PSC1_RX,
 		.rx_ipr = BCOM_IPR_PSC1_RX,
Signature · 1 line
2.54.0
#1 lore raw
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] Inadequate or missing bounds checking on psc_num can lead to out-of-bounds array access in bcom_psc_params. --

commit c24864efcf21b2526be68d727e70076f6cee258c
Author: Rosen Penev <[email protected]>

dma: bestcomm: gen_bd: split struct bcom_psc_params from array definition

This commit splits the bcom_psc_params struct definition from its array declaration to resolve kernel-doc parser warnings. It also adds missing members and marks the lookup table as const so it can be moved to read only memory.

Quoted 5 lines
drivers/dma/bestcomm/gen_bd.c +0 -0
index 8a24a5cbc2633..61b5746e1a972 100644
@@ -254,17 +254,23 @@ EXPORT_SYMBOL_GPL(bcom_gen_bd_tx_release);

[ ... ]

drivers/dma/bestcomm/gen_bd.c EXPORT_SYMBOL_GPL(bcom_gen_bd_tx_release); · line 273

Quoted 10 lines
drivers/dma/bestcomm/gen_bd.c +4 -2
-static struct bcom_psc_params {
+struct bcom_psc_params {
 	int rx_initiator;
 	int rx_ipr;
 	int tx_initiator;
 	int tx_ipr;
-} bcom_psc_params[] = {
+};
+
+static const struct bcom_psc_params bcom_psc_params[] = {

[Severity: High]
This is a pre-existing issue, but does this fixed-size array expose an out-of-bounds access?

The array strictly contains 6 elements (indices 0 through 5). However, bcom_psc_gen_bd_rx_init() checks the psc_num index against MPC52xx_PSC_MAXNUM:

drivers/dma/bestcomm/gen_bd.c:bcom_psc_gen_bd_rx_init() {
        if (psc_num >= MPC52xx_PSC_MAXNUM)
                return NULL;

        ...
        bcom_psc_params[psc_num].rx_initiator,
        ...
}

If CONFIG_PPC_MPC512x is enabled, MPC52xx_PSC_MAXNUM can be 12, allowing an index up to 11 to access the 6-element bcom_psc_params[].

Additionally, bcom_psc_gen_bd_tx_init() appears to perform no bounds checking on psc_num at all before indexing this array. Callers like mpc5200_audio_dma_create() can pass an unvalidated cell-index from the device tree directly into these functions.

Can a malformed device tree trigger out-of-bounds memory reads here, leading to out-of-bounds MMIO writes when using the read garbage as the initiator and ipr?

Quoted 2 lines
drivers/dma/bestcomm/gen_bd.c +0 -0
 	[0] = {
 		.rx_initiator = BCOM_INITIATOR_PSC1_RX,
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1
#2 lore raw
Quoted 56 lines · On Fri, May 29, 2026 at 7:19 PM <[email protected]> wrote:

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] Inadequate or missing bounds checking on psc_num can lead to out-of-bounds array access in bcom_psc_params. --

commit c24864efcf21b2526be68d727e70076f6cee258c
Author: Rosen Penev <[email protected]>

dma: bestcomm: gen_bd: split struct bcom_psc_params from array definition

This commit splits the bcom_psc_params struct definition from its array declaration to resolve kernel-doc parser warnings. It also adds missing members and marks the lookup table as const so it can be moved to read only memory.

Quoted 5 lines
drivers/dma/bestcomm/gen_bd.c +0 -0
index 8a24a5cbc2633..61b5746e1a972 100644
@@ -254,17 +254,23 @@ EXPORT_SYMBOL_GPL(bcom_gen_bd_tx_release);

[ ... ]

drivers/dma/bestcomm/gen_bd.c EXPORT_SYMBOL_GPL(bcom_gen_bd_tx_release); · line 273

Quoted 10 lines
drivers/dma/bestcomm/gen_bd.c +4 -2
-static struct bcom_psc_params {
+struct bcom_psc_params {
      int rx_initiator;
      int rx_ipr;
      int tx_initiator;
      int tx_ipr;
-} bcom_psc_params[] = {
+};
+
+static const struct bcom_psc_params bcom_psc_params[] = {

[Severity: High]
This is a pre-existing issue, but does this fixed-size array expose an out-of-bounds access?

The array strictly contains 6 elements (indices 0 through 5). However, bcom_psc_gen_bd_rx_init() checks the psc_num index against MPC52xx_PSC_MAXNUM:

drivers/dma/bestcomm/gen_bd.c:bcom_psc_gen_bd_rx_init() {
        if (psc_num >= MPC52xx_PSC_MAXNUM)
                return NULL;

        ...
        bcom_psc_params[psc_num].rx_initiator,
        ...
}

If CONFIG_PPC_MPC512x is enabled, MPC52xx_PSC_MAXNUM can be 12, allowing an index up to 11 to access the 6-element bcom_psc_params[].

Quoted 8 lines

Additionally, bcom_psc_gen_bd_tx_init() appears to perform no bounds checking on psc_num at all before indexing this array. Callers like mpc5200_audio_dma_create() can pass an unvalidated cell-index from the device tree directly into these functions.

Can a malformed device tree trigger out-of-bounds memory reads here, leading to out-of-bounds MMIO writes when using the read garbage as the initiator and ipr?

Fixed in different patch.

Quoted 6 lines
Quoted 3 lines

[0] = {
.rx_initiator = BCOM_INITIATOR_PSC1_RX,

--
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1