Re: [PATCH 06/12] libata: separate out ata_acpi_gtm_xfermask() from pacpi_discover_modes()

[Tejun Heo <htejun@gmail.com>]

Alan Cox wrote:
>> +/* Welcome to ACPI, bring a bucket */
>> +const unsigned int ata_acpi_pio_cycle[7] = {
>> +	600, 383, 240, 180, 120, 100, 80
>> +};
>> +EXPORT_SYMBOL_GPL(ata_acpi_pio_cycle);
>> +
>> +const unsigned int ata_acpi_mwdma_cycle[5] = {
>> +	480, 150, 120, 100, 80
>> +};
>> +EXPORT_SYMBOL_GPL(ata_acpi_mwdma_cycle);
>> +
>> +const unsigned int ata_acpi_udma_cycle[7] = {
>> +	120, 80, 60, 45, 30, 20, 15
>> +};
>> +EXPORT_SYMBOL_GPL(ata_acpi_udma_cycle);
> 
> Do we really need to keep exporting all these things. So far this patch
> set has exported a set of very specific ACPI arrays and a load of
> internal functions. That to me says the splitting up is wrong.
> 
> One option would be to make those tables private and simply make the
> pata_acpi driver use the ata_timing functions

This is mid-step of merging ACPI timing handling into the standard
ata_timing mechanism.  These will go away in later patch.

-- 
tejun