Re: [PATCH 5/5] Bluetooth: btmrvl: Use common error handling code in btmrvl_sdio_download_fw_w_helper()

[SF Markus Elfring <elfring@users.sourceforge.net>]

>> Add a jump target so that the setting of a specific error code is stored
>> only once at the end of this function.
>>
>> Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
>> ---
>> drivers/bluetooth/btmrvl_sdio.c | 13 +++++++------
>> 1 file changed, 7 insertions(+), 6 deletions(-)
>>
>> diff --git a/drivers/bluetooth/btmrvl_sdio.c b/drivers/bluetooth/btmrvl_sdio.c
>> index 05c78fcc13ff..24ed62fe2aeb 100644
>> --- a/drivers/bluetooth/btmrvl_sdio.c
>> +++ b/drivers/bluetooth/btmrvl_sdio.c
>> @@ -601,8 +601,7 @@ static int btmrvl_sdio_download_fw_w_helper(struct btmrvl_sdio_card *card)
>> 					" base0 = 0x%04X(%d)."
>> 					" Terminating download",
>> 					base0, base0);
>> -				ret = -EIO;
>> -				goto done;
>> +				goto e_io;
>> 			}
>> 			base1 = sdio_readb(card->func,
>> 					card->reg->sq_read_base_addr_a1, &ret);
>> @@ -611,8 +610,7 @@ static int btmrvl_sdio_download_fw_w_helper(struct btmrvl_sdio_card *card)
>> 					" base1 = 0x%04X(%d)."
>> 					" Terminating download",
>> 					base1, base1);
>> -				ret = -EIO;
>> -				goto done;
>> +				goto e_io;
>> 			}
>>
>> 			len = (((u16) base1) << 8) | base0;
>> @@ -638,8 +636,7 @@ static int btmrvl_sdio_download_fw_w_helper(struct btmrvl_sdio_card *card)
>> 			if (count > MAX_WRITE_IOMEM_RETRY) {
>> 				BT_ERR("FW download failure @%d, "
>> 					"over max retry count", offset);
>> -				ret = -EIO;
>> -				goto done;
>> +				goto e_io;
>> 			}
>> 			BT_ERR("FW CRC error indicated by the helper: "
>> 				"len = 0x%04X, txlen = %d", len, txlen);
>> @@ -681,6 +678,10 @@ static int btmrvl_sdio_download_fw_w_helper(struct btmrvl_sdio_card *card)
>> 	kfree(tmpfwbuf);
>> 	release_firmware(fw_firmware);
>> 	return ret;
>> +
>> +e_io:
>> +	ret = -EIO;
>> +	goto done;
>> }
> 
> I am not applying this one. I see zero benefit in this change.

Would you care for a bit of object code reduction in this function implementation.


> It is not even saving a single line since it actually is more code.

Should I fiddle with any other source code layout?

Do you find an extra blank line inappropriate before the added jump target
in this use case?

Regards,
Markus