[AMD Official Use Only - AMD Internal Distribution Only]
The limit variable in amdgpu_eeprom_xfer is not 0, so the buf_size will be cut into smaller pieces, u16 is enough for __amdgpu_eeprom_xfer.
Anyway, use u32 for __amdgpu_eeprom_xfer and make sure the msgs[1].len is less than U16_MAX is better, will create a new patch for the purpose.
Tao
_____________________________________________
From: Zhang, Hawking <Hawking.Zhang@xxxxxxx>
Sent: Monday, May 20, 2024 11:23 AM
To: Zhou1, Tao <Tao.Zhou1@xxxxxxx>; amd-gfx@xxxxxxxxxxxxxxxxxxxxx
Cc: Zhou1, Tao <Tao.Zhou1@xxxxxxx>
Subject: RE: [PATCH] drm/amdgpu: update type of buf size to u32 for eeprom functions
From: Zhang, Hawking <Hawking.Zhang@xxxxxxx>
Sent: Monday, May 20, 2024 11:23 AM
To: Zhou1, Tao <Tao.Zhou1@xxxxxxx>; amd-gfx@xxxxxxxxxxxxxxxxxxxxx
Cc: Zhou1, Tao <Tao.Zhou1@xxxxxxx>
Subject: RE: [PATCH] drm/amdgpu: update type of buf size to u32 for eeprom functions
[AMD Official Use Only - AMD Internal Distribution Only]
Hmm… but in __amdgpu_eeprom_xfer, the u32 will still be cut to u16.
__amdgpu_eeprom_xfer(struct i2c_adapter *i2c_adap, u32 eeprom_addr, u8 *eeprom_buf, u16 buf_size, bool read)
Regards,
Hawking
Hawking
-----Original Message-----
From: amd-gfx <amd-gfx-bounces@xxxxxxxxxxxxxxxxxxxxx> On Behalf Of Tao Zhou
Sent: Monday, May 20, 2024 10:46
To: amd-gfx@xxxxxxxxxxxxxxxxxxxxx
Cc: Zhou1, Tao <Tao.Zhou1@xxxxxxx>
Subject: [PATCH] drm/amdgpu: update type of buf size to u32 for eeprom functions
From: amd-gfx <amd-gfx-bounces@xxxxxxxxxxxxxxxxxxxxx> On Behalf Of Tao Zhou
Sent: Monday, May 20, 2024 10:46
To: amd-gfx@xxxxxxxxxxxxxxxxxxxxx
Cc: Zhou1, Tao <Tao.Zhou1@xxxxxxx>
Subject: [PATCH] drm/amdgpu: update type of buf size to u32 for eeprom functions
Avoid overflow issue.
Signed-off-by: Tao Zhou <tao.zhou1@xxxxxxx>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.c | 6 +++--- drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.h | 4 ++--
2 files changed, 5 insertions(+), 5 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.c
index e71768661ca8..09a34c7258e2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.c
@@ -179,7 +179,7 @@ static int __amdgpu_eeprom_xfer(struct i2c_adapter *i2c_adap, u32 eeprom_addr,
* Returns the number of bytes read/written; -errno on error.
*/
static int amdgpu_eeprom_xfer(struct i2c_adapter *i2c_adap, u32 eeprom_addr,
- u8 *eeprom_buf, u16 buf_size, bool read)
+ u8 *eeprom_buf, u32 buf_size, bool read)
{
const struct i2c_adapter_quirks *quirks = i2c_adap->quirks;
u16 limit;
@@ -225,7 +225,7 @@ static int amdgpu_eeprom_xfer(struct i2c_adapter *i2c_adap, u32 eeprom_addr,
int amdgpu_eeprom_read(struct i2c_adapter *i2c_adap,
u32 eeprom_addr, u8 *eeprom_buf,
- u16 bytes)
+ u32 bytes)
{
return amdgpu_eeprom_xfer(i2c_adap, eeprom_addr, eeprom_buf, bytes,
true);
@@ -233,7 +233,7 @@ int amdgpu_eeprom_read(struct i2c_adapter *i2c_adap,
int amdgpu_eeprom_write(struct i2c_adapter *i2c_adap,
u32 eeprom_addr, u8 *eeprom_buf,
- u16 bytes)
+ u32 bytes)
{
return amdgpu_eeprom_xfer(i2c_adap, eeprom_addr, eeprom_buf, bytes,
false);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.h
index 6935adb2be1f..8083b8253ef4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.h
@@ -28,10 +28,10 @@
int amdgpu_eeprom_read(struct i2c_adapter *i2c_adap,
u32 eeprom_addr, u8 *eeprom_buf,
- u16 bytes);
+ u32 bytes);
int amdgpu_eeprom_write(struct i2c_adapter *i2c_adap,
u32 eeprom_addr, u8 *eeprom_buf,
- u16 bytes);
+ u32 bytes);
#endif
--
2.34.1
