[PATCH v1] Bluetooth: btintel_pcie: Align shared DMA memory to 128 bytes

[PATCH v1] Bluetooth: btintel_pcie: Align shared DMA memory to 128 bytes

[Kiran K]

Align each descriptor/index/context region to 128 bytes before
calculating the total DMA pool size. This ensures the memory layout
shared with firmware meets the 128-byte alignment requirement.

The DMA pool alignment is also set to 128 bytes to match the firmware
expectation for all shared structures.

Signed-off-by: Kiran K <kiran.k@intel.com>
---
 drivers/bluetooth/btintel_pcie.c | 94 ++++++++++++++++++--------------
 drivers/bluetooth/btintel_pcie.h |  3 -
 2 files changed, 53 insertions(+), 44 deletions(-)

diff --git a/drivers/bluetooth/btintel_pcie.c b/drivers/bluetooth/btintel_pcie.c
index bd1c6116715e..c4c19a25157a 100644
--- a/drivers/bluetooth/btintel_pcie.c
+++ b/drivers/bluetooth/btintel_pcie.c
@@ -37,6 +37,8 @@
 
 #define POLL_INTERVAL_US	10
 
+#define BTINTEL_PCIE_DMA_ALIGN_128B	128 /* 128 byte aligned */
+
 /* Intel Bluetooth PCIe device id table */
 static const struct pci_device_id btintel_pcie_table[] = {
 	/* BlazarI, Wildcat Lake */
@@ -1756,27 +1758,6 @@ static int btintel_pcie_setup_rxq_bufs(struct btintel_pcie_data *data,
 	return 0;
 }
 
-static void btintel_pcie_setup_ia(struct btintel_pcie_data *data,
-				  dma_addr_t p_addr, void *v_addr,
-				  struct ia *ia)
-{
-	/* TR Head Index Array */
-	ia->tr_hia_p_addr = p_addr;
-	ia->tr_hia = v_addr;
-
-	/* TR Tail Index Array */
-	ia->tr_tia_p_addr = p_addr + sizeof(u16) * BTINTEL_PCIE_NUM_QUEUES;
-	ia->tr_tia = v_addr + sizeof(u16) * BTINTEL_PCIE_NUM_QUEUES;
-
-	/* CR Head index Array */
-	ia->cr_hia_p_addr = p_addr + (sizeof(u16) * BTINTEL_PCIE_NUM_QUEUES * 2);
-	ia->cr_hia = v_addr + (sizeof(u16) * BTINTEL_PCIE_NUM_QUEUES * 2);
-
-	/* CR Tail Index Array */
-	ia->cr_tia_p_addr = p_addr + (sizeof(u16) * BTINTEL_PCIE_NUM_QUEUES * 3);
-	ia->cr_tia = v_addr + (sizeof(u16) * BTINTEL_PCIE_NUM_QUEUES * 3);
-}
-
 static void btintel_pcie_free(struct btintel_pcie_data *data)
 {
 	btintel_pcie_free_rxq_bufs(data, &data->rxq);
@@ -1794,13 +1775,16 @@ static int btintel_pcie_alloc(struct btintel_pcie_data *data)
 	size_t total;
 	dma_addr_t p_addr;
 	void *v_addr;
+	size_t tfd_size, frbd_size, ctx_size, ci_size, urbd0_size, urbd1_size;
 
 	/* Allocate the chunk of DMA memory for descriptors, index array, and
 	 * context information, instead of allocating individually.
 	 * The DMA memory for data buffer is allocated while setting up the
 	 * each queue.
 	 *
-	 * Total size is sum of the following
+	 * Total size is sum of the following and each of the individual sizes
+	 * are aligned to 128 bytes before adding up.
+	 *
 	 *  + size of TFD * Number of descriptors in queue
 	 *  + size of URBD0 * Number of descriptors in queue
 	 *  + size of FRBD * Number of descriptors in queue
@@ -1808,15 +1792,25 @@ static int btintel_pcie_alloc(struct btintel_pcie_data *data)
 	 *  + size of index * Number of queues(2) * type of index array(4)
 	 *  + size of context information
 	 */
-	total = (sizeof(struct tfd) + sizeof(struct urbd0)) * BTINTEL_PCIE_TX_DESCS_COUNT;
-	total += (sizeof(struct frbd) + sizeof(struct urbd1)) * BTINTEL_PCIE_RX_DESCS_COUNT;
+	tfd_size = ALIGN(sizeof(struct tfd) * BTINTEL_PCIE_TX_DESCS_COUNT,
+			 BTINTEL_PCIE_DMA_ALIGN_128B);
+	urbd0_size = ALIGN(sizeof(struct urbd0) * BTINTEL_PCIE_TX_DESCS_COUNT,
+			   BTINTEL_PCIE_DMA_ALIGN_128B);
 
-	/* Add the sum of size of index array and size of ci struct */
-	total += (sizeof(u16) * BTINTEL_PCIE_NUM_QUEUES * 4) + sizeof(struct ctx_info);
+	frbd_size = ALIGN(sizeof(struct frbd) * BTINTEL_PCIE_RX_DESCS_COUNT,
+			  BTINTEL_PCIE_DMA_ALIGN_128B);
+	urbd1_size = ALIGN(sizeof(struct urbd1) * BTINTEL_PCIE_RX_DESCS_COUNT,
+			   BTINTEL_PCIE_DMA_ALIGN_128B);
+
+	ci_size = ALIGN(sizeof(u16) * BTINTEL_PCIE_NUM_QUEUES,
+			BTINTEL_PCIE_DMA_ALIGN_128B);
+
+	ctx_size = ALIGN(sizeof(struct ctx_info), BTINTEL_PCIE_DMA_ALIGN_128B);
+
+	total = tfd_size + urbd0_size + frbd_size + urbd1_size + ctx_size + ci_size * 4;
 
-	/* Allocate DMA Pool */
 	data->dma_pool = dma_pool_create(KBUILD_MODNAME, &data->pdev->dev,
-					 total, BTINTEL_PCIE_DMA_POOL_ALIGNMENT, 0);
+					 total, BTINTEL_PCIE_DMA_ALIGN_128B, 0);
 	if (!data->dma_pool) {
 		err = -ENOMEM;
 		goto exit_error;
@@ -1841,29 +1835,29 @@ static int btintel_pcie_alloc(struct btintel_pcie_data *data)
 	data->txq.tfds_p_addr = p_addr;
 	data->txq.tfds = v_addr;
 
-	p_addr += (sizeof(struct tfd) * BTINTEL_PCIE_TX_DESCS_COUNT);
-	v_addr += (sizeof(struct tfd) * BTINTEL_PCIE_TX_DESCS_COUNT);
+	p_addr += tfd_size;
+	v_addr += tfd_size;
 
 	/* Setup urbd0 */
 	data->txq.urbd0s_p_addr = p_addr;
 	data->txq.urbd0s = v_addr;
 
-	p_addr += (sizeof(struct urbd0) * BTINTEL_PCIE_TX_DESCS_COUNT);
-	v_addr += (sizeof(struct urbd0) * BTINTEL_PCIE_TX_DESCS_COUNT);
+	p_addr += urbd0_size;
+	v_addr += urbd0_size;
 
 	/* Setup FRBD*/
 	data->rxq.frbds_p_addr = p_addr;
 	data->rxq.frbds = v_addr;
 
-	p_addr += (sizeof(struct frbd) * BTINTEL_PCIE_RX_DESCS_COUNT);
-	v_addr += (sizeof(struct frbd) * BTINTEL_PCIE_RX_DESCS_COUNT);
+	p_addr += frbd_size;
+	v_addr += frbd_size;
 
 	/* Setup urbd1 */
 	data->rxq.urbd1s_p_addr = p_addr;
 	data->rxq.urbd1s = v_addr;
 
-	p_addr += (sizeof(struct urbd1) * BTINTEL_PCIE_RX_DESCS_COUNT);
-	v_addr += (sizeof(struct urbd1) * BTINTEL_PCIE_RX_DESCS_COUNT);
+	p_addr += urbd1_size;
+	v_addr += urbd1_size;
 
 	/* Setup data buffers for txq */
 	err = btintel_pcie_setup_txq_bufs(data, &data->txq);
@@ -1875,8 +1869,29 @@ static int btintel_pcie_alloc(struct btintel_pcie_data *data)
 	if (err)
 		goto exit_error_txq;
 
-	/* Setup Index Array */
-	btintel_pcie_setup_ia(data, p_addr, v_addr, &data->ia);
+	/* TR Head Index Array */
+	data->ia.tr_hia_p_addr = p_addr;
+	data->ia.tr_hia = v_addr;
+	p_addr += ci_size;
+	v_addr += ci_size;
+
+	/* TR Tail Index Array */
+	data->ia.tr_tia_p_addr = p_addr;
+	data->ia.tr_tia = v_addr;
+	p_addr += ci_size;
+	v_addr += ci_size;
+
+	/* CR Head index Array */
+	data->ia.cr_hia_p_addr = p_addr;
+	data->ia.cr_hia = v_addr;
+	p_addr += ci_size;
+	v_addr += ci_size;
+
+	/* CR Tail Index Array */
+	data->ia.cr_tia_p_addr = p_addr;
+	data->ia.cr_tia = v_addr;
+	p_addr += ci_size;
+	v_addr += ci_size;
 
 	/* Setup data buffers for dbgc */
 	err = btintel_pcie_setup_dbgc(data);
@@ -1884,9 +1899,6 @@ static int btintel_pcie_alloc(struct btintel_pcie_data *data)
 		goto exit_error_txq;
 
 	/* Setup Context Information */
-	p_addr += sizeof(u16) * BTINTEL_PCIE_NUM_QUEUES * 4;
-	v_addr += sizeof(u16) * BTINTEL_PCIE_NUM_QUEUES * 4;
-
 	data->ci = v_addr;
 	data->ci_p_addr = p_addr;
 
diff --git a/drivers/bluetooth/btintel_pcie.h b/drivers/bluetooth/btintel_pcie.h
index e3d941ffef4a..3c7bb708362d 100644
--- a/drivers/bluetooth/btintel_pcie.h
+++ b/drivers/bluetooth/btintel_pcie.h
@@ -178,9 +178,6 @@ enum {
 /* The size of DMA buffer for TX and RX in bytes */
 #define BTINTEL_PCIE_BUFFER_SIZE	4096
 
-/* DMA allocation alignment */
-#define BTINTEL_PCIE_DMA_POOL_ALIGNMENT	256
-
 #define BTINTEL_PCIE_TX_WAIT_TIMEOUT_MS		500
 
 /* Doorbell vector for TFD */
-- 
2.43.0

RE: [v1] Bluetooth: btintel_pcie: Align shared DMA memory to 128 bytes

[bluez.test.bot]

[-- Attachment #1: Type: text/plain, Size: 2942 bytes --]

This is automated email and please do not reply to this email!

Dear submitter,

Thank you for submitting the patches to the linux bluetooth mailing list.
This is a CI test results with your patch series:
PW Link:https://patchwork.kernel.org/project/bluetooth/list/?series=1060579

---Test result---

Test Summary:
CheckPatch                    PENDING   0.38 seconds
GitLint                       PENDING   0.26 seconds
SubjectPrefix                 PASS      0.09 seconds
BuildKernel                   PASS      26.75 seconds
CheckAllWarning               PASS      29.25 seconds
CheckSparse                   PASS      33.32 seconds
BuildKernel32                 PASS      26.52 seconds
TestRunnerSetup               PASS      580.39 seconds
TestRunner_l2cap-tester       FAIL      32.61 seconds
TestRunner_iso-tester         PASS      94.50 seconds
TestRunner_bnep-tester        PASS      6.38 seconds
TestRunner_mgmt-tester        FAIL      116.42 seconds
TestRunner_rfcomm-tester      PASS      10.16 seconds
TestRunner_sco-tester         FAIL      14.92 seconds
TestRunner_ioctl-tester       PASS      10.55 seconds
TestRunner_mesh-tester        FAIL      12.45 seconds
TestRunner_smp-tester         PASS      8.86 seconds
TestRunner_userchan-tester    PASS      6.83 seconds
IncrementalBuild              PENDING   0.99 seconds

Details
##############################
Test: CheckPatch - PENDING
Desc: Run checkpatch.pl script
Output:

##############################
Test: GitLint - PENDING
Desc: Run gitlint
Output:

##############################
Test: TestRunner_l2cap-tester - FAIL
Desc: Run l2cap-tester with test-runner
Output:
Total: 96, Passed: 94 (97.9%), Failed: 2, Not Run: 0

Failed Test Cases
L2CAP LE Client - Read 32k Success                   Timed out    2.406 seconds
L2CAP LE Client - RX Timestamping 32k                Timed out    1.888 seconds
##############################
Test: TestRunner_mgmt-tester - FAIL
Desc: Run mgmt-tester with test-runner
Output:
Total: 494, Passed: 489 (99.0%), Failed: 1, Not Run: 4

Failed Test Cases
Read Exp Feature - Success                           Failed       0.114 seconds
##############################
Test: TestRunner_sco-tester - FAIL
Desc: Run sco-tester with test-runner
Output:
WARNING: possible circular locking dependency detected
BUG: sleeping function called from invalid context at net/core/sock.c:3782
Total: 30, Passed: 30 (100.0%), Failed: 0, Not Run: 0
##############################
Test: TestRunner_mesh-tester - FAIL
Desc: Run mesh-tester with test-runner
Output:
Total: 10, Passed: 8 (80.0%), Failed: 2, Not Run: 0

Failed Test Cases
Mesh - Send cancel - 1                               Timed out    2.773 seconds
Mesh - Send cancel - 2                               Timed out    1.991 seconds
##############################
Test: IncrementalBuild - PENDING
Desc: Incremental build with the patches in the series
Output:



---
Regards,
Linux Bluetooth