svn commit: r188918 - head/sys/dev/ata/chipsets

[Alexander Motin]

Gavin Atkinson wrote:
> On Sun, 22 Feb 2009, Alexander Motin wrote:
> 
>> Author: mav
>> Date: Sun Feb 22 14:08:20 2009
>> New Revision: 188918
>> URL: http://svn.freebsd.org/changeset/base/188918
>>
>> Log:
>>  Tune AHCI verbose messages to make log more readable.
>>
> [snip]
>> @@ -665,9 +670,10 @@ ata_ahci_softreset(device_t dev, int por
>>     //ctp->cfis[7] = ATA_D_LBA | ATA_D_IBM;
>>     ctp->cfis[15] = (ATA_A_4BIT | ATA_A_RESET);
>>
>> -    if (ata_ahci_issue_cmd(dev, ATA_AHCI_CMD_RESET |
>> ATA_AHCI_CMD_CLR_BUSY,100))
>> -    device_printf(dev, "setting SRST failed ??\n");
>> -    //return -1;
>> +    if (ata_ahci_issue_cmd(dev, ATA_AHCI_CMD_RESET |
>> ATA_AHCI_CMD_CLR_BUSY,100)) {
>> +    device_printf(dev, "software reset set timeout\n");
>> +    return (-1);
>> +    }
>>
> 
> Was the uncommenting of the return intentional here?  Before, the code
> would continue to clear the soft reset even if it didn't appear to have
> worked.

It was. It shouldn't happen now. Linux does the same.

>> @@ -767,11 +778,11 @@ ata_ahci_reset(device_t dev)
>>     break;
>>     default: /* SOS XXX */
>>     if (bootverbose)
>> -        device_printf(dev, "No signature, asuming disk device\n");
>> +        device_printf(dev, "Unknown signature, asuming disk device\n");
>>     ch->devices = ATA_ATA_MASTER;
>>     }
> 
> Spelling is wrong here (and was before you changed it too): should be
> "assuming" not "asuming"

Thanks! Fixed.

-- 
Alexander Motin