[PATCH v3] bus: mhi: host: fix endianness of BHI vector table

[PATCH v3] bus: mhi: host: fix endianness of BHI vector table

[Alexander Wilhelm]

On big endian platform like PowerPC the MHI bus does not start properly. The
following example shows the error messages by using qcn9274 wireless radio
module with ath12k driver:

    ath12k_pci 0001:01:00.0: BAR 0: assigned [mem 0xc00000000-0xc001fffff 64bit]
    ath12k_pci 0001:01:00.0: MSI vectors: 1
    ath12k_pci 0001:01:00.0: Hardware name: qcn9274 hw2.0
    ath12k_pci 0001:01:00.0: failed to set mhi state: POWER_ON(2)
    ath12k_pci 0001:01:00.0: failed to start mhi: -110
    ath12k_pci 0001:01:00.0: failed to power up :-110
    ath12k_pci 0001:01:00.0: failed to create soc core: -110
    ath12k_pci 0001:01:00.0: failed to init core: -110
    ath12k_pci: probe of 0001:01:00.0 failed with error -110

Fix it by swapping DMA address and size of the BHI vector table.

Signed-off-by: Alexander Wilhelm <alexander.wilhelm@westermo.com>
Reviewed-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com>
---
Changes in v2:
  - Set __le64 type for dma_addr and size.

Changes in v3:
  - Improve the clarity of the commit message.

 drivers/bus/mhi/host/boot.c     | 8 ++++----
 drivers/bus/mhi/host/internal.h | 4 ++--
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/bus/mhi/host/boot.c b/drivers/bus/mhi/host/boot.c
index efa3b6dddf4d..205d83ac069f 100644
--- a/drivers/bus/mhi/host/boot.c
+++ b/drivers/bus/mhi/host/boot.c
@@ -31,8 +31,8 @@ int mhi_rddm_prepare(struct mhi_controller *mhi_cntrl,
 	int ret;
 
 	for (i = 0; i < img_info->entries - 1; i++, mhi_buf++, bhi_vec++) {
-		bhi_vec->dma_addr = mhi_buf->dma_addr;
-		bhi_vec->size = mhi_buf->len;
+		bhi_vec->dma_addr = cpu_to_le64(mhi_buf->dma_addr);
+		bhi_vec->size = cpu_to_le64(mhi_buf->len);
 	}
 
 	dev_dbg(dev, "BHIe programming for RDDM\n");
@@ -431,8 +431,8 @@ static void mhi_firmware_copy_bhie(struct mhi_controller *mhi_cntrl,
 	while (remainder) {
 		to_cpy = min(remainder, mhi_buf->len);
 		memcpy(mhi_buf->buf, buf, to_cpy);
-		bhi_vec->dma_addr = mhi_buf->dma_addr;
-		bhi_vec->size = to_cpy;
+		bhi_vec->dma_addr = cpu_to_le64(mhi_buf->dma_addr);
+		bhi_vec->size = cpu_to_le64(to_cpy);
 
 		buf += to_cpy;
 		remainder -= to_cpy;
diff --git a/drivers/bus/mhi/host/internal.h b/drivers/bus/mhi/host/internal.h
index ce566f7d2e92..1dbc3f736161 100644
--- a/drivers/bus/mhi/host/internal.h
+++ b/drivers/bus/mhi/host/internal.h
@@ -25,8 +25,8 @@ struct mhi_ctxt {
 };
 
 struct bhi_vec_entry {
-	u64 dma_addr;
-	u64 size;
+	__le64 dma_addr;
+	__le64 size;
 };
 
 enum mhi_fw_load_type {
-- 
2.34.1

Re: [PATCH v3] bus: mhi: host: fix endianness of BHI vector table

[Jeff Hugo]

On 5/19/2025 12:10 AM, Alexander Wilhelm wrote:
> On big endian platform like PowerPC the MHI bus does not start properly. The
> following example shows the error messages by using qcn9274 wireless radio
> module with ath12k driver:
> 
>      ath12k_pci 0001:01:00.0: BAR 0: assigned [mem 0xc00000000-0xc001fffff 64bit]
>      ath12k_pci 0001:01:00.0: MSI vectors: 1
>      ath12k_pci 0001:01:00.0: Hardware name: qcn9274 hw2.0
>      ath12k_pci 0001:01:00.0: failed to set mhi state: POWER_ON(2)
>      ath12k_pci 0001:01:00.0: failed to start mhi: -110
>      ath12k_pci 0001:01:00.0: failed to power up :-110
>      ath12k_pci 0001:01:00.0: failed to create soc core: -110
>      ath12k_pci 0001:01:00.0: failed to init core: -110
>      ath12k_pci: probe of 0001:01:00.0 failed with error -110
> 
> Fix it by swapping DMA address and size of the BHI vector table.
> 
> Signed-off-by: Alexander Wilhelm <alexander.wilhelm@westermo.com>
> Reviewed-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com>

Shouldn't there be a fixes tag?  Probably pointing to one of the 
original MHI commits given what this change is touching.

-Jeff

Re: [PATCH v3] bus: mhi: host: fix endianness of BHI vector table

[Alexander Wilhelm]

Am Mon, May 19, 2025 at 08:13:00AM -0600 schrieb Jeff Hugo:
> On 5/19/2025 12:10 AM, Alexander Wilhelm wrote:
> > On big endian platform like PowerPC the MHI bus does not start properly. The
> > following example shows the error messages by using qcn9274 wireless radio
> > module with ath12k driver:
> > 
> >      ath12k_pci 0001:01:00.0: BAR 0: assigned [mem 0xc00000000-0xc001fffff 64bit]
> >      ath12k_pci 0001:01:00.0: MSI vectors: 1
> >      ath12k_pci 0001:01:00.0: Hardware name: qcn9274 hw2.0
> >      ath12k_pci 0001:01:00.0: failed to set mhi state: POWER_ON(2)
> >      ath12k_pci 0001:01:00.0: failed to start mhi: -110
> >      ath12k_pci 0001:01:00.0: failed to power up :-110
> >      ath12k_pci 0001:01:00.0: failed to create soc core: -110
> >      ath12k_pci 0001:01:00.0: failed to init core: -110
> >      ath12k_pci: probe of 0001:01:00.0 failed with error -110
> > 
> > Fix it by swapping DMA address and size of the BHI vector table.
> > 
> > Signed-off-by: Alexander Wilhelm <alexander.wilhelm@westermo.com>
> > Reviewed-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com>
> 
> Shouldn't there be a fixes tag?  Probably pointing to one of the original
> MHI commits given what this change is touching.
> 
> -Jeff

Thank you for supporting me. I will add that in the next patch version. This
patch fixes 3 different origin commits. I would take the first one, where
bhi_vec_entry was first introducted. The following 2 commits are only follow up
errors.

Best regards
Alexander Wilhelm