clean up kernel_{read,write} & friends v2

[Al Viro]

On Thu, May 28, 2020 at 12:44:41PM -0700, Matthew Wilcox wrote:
> On Thu, May 28, 2020 at 08:33:40PM +0100, Al Viro wrote:
> > On Thu, May 28, 2020 at 12:22:08PM -0700, Joe Perches wrote:
> > 
> > > Hard limits at 80 really don't work well, especially with
> > > some of the 25+ character length identifiers used today.
> > 
> > IMO any such identifier is a good reason for a warning.
> > 
> > The litmus test is actually very simple: how unpleasant would it be
> > to mention the identifiers while discussing the code over the phone?
> 
> Here's a good example of a function which should be taken out and shot:
> 
> int amdgpu_atombios_get_leakage_vddc_based_on_leakage_params(struct amdgpu_devic
> e *adev,
> ...
>         switch (frev) {
>         case 2:
>                 switch (crev) {
> ...
>                         if (profile->ucElbVDDC_Num > 0) {
>                                 for (i = 0; i < profile->ucElbVDDC_Num; i++) {
>                                         if (vddc_id_buf[i] == virtual_voltage_id) {
>                                                 for (j = 0; j < profile->ucLeakageBinNum; j++) {
>                                                         if (vbios_voltage_id <= leakage_bin[j]) {
>                                                                 *vddc = vddc_buf[j * profile->ucElbVDDC_Num + i];
> 
> I mean, I get it that AMD want to show off just how studly a monitor they
> support, but good grief ...

I must respectfully disagree.  It clearly needs to be hanged, drawn and
quartered...