riscv: Add macro for hfence.gvma instruction

[doonbsd_gmail.com]

Summary

Add an inline function hfence_gvma() for hfence.gvma instruction.

Signed-off-by: Doongar Singh <[e-mail]>

[markj]

sys/riscv/include/atomic.h

41 ↗

(On Diff #170325)

Let's put it in riscv/include/cpufunc.h instead? atomic.h isn't really the right place for this, it's more similar to sfence_vma() etc..

[doonbsd_gmail.com]

Agreed. I will update the patch.

[doonbsd_gmail.com]

Moved hfence_gvma from atomic.h to cpufunc.h and changed it to an inline function.

[doonbsd_gmail.com]

riscv: Add an inline function hfence_gvma() for hfence.gvma instruction.

[doonbsd_gmail.com]

Hold on. I just found one style problem. There must be a empty line after start curly brace when the function has single line body.

I send an updated patch.

[jrtc27]

In D54857#1254516, @doonbsd_gmail.com wrote:

Hold on. I just found one style problem. There must be a empty line after start curly brace when the function has single line body.

I send an updated patch.

That style(9) rule was dropped years ago

[doonbsd_gmail.com]

So, this can't be merged now? Shall I drop that change and update new diff?

[markj]

In D54857#1254686, @doonbsd_gmail.com wrote:

So, this can't be merged now? Shall I drop that change and update new diff?

No, it's ok as-is. Indeed, style(9) doesn't require the blank line, but most of the other wrappers in this file have one so we might as well stay consistent.

[doonbsd_gmail.com]

In D54857#1254906, @markj wrote:

In D54857#1254686, @doonbsd_gmail.com wrote:

So, this can't be merged now? Shall I drop that change and update new diff?

No, it's ok as-is. Indeed, style(9) doesn't require the blank line, but most of the other wrappers in this file have one so we might as well stay consistent.

Ok. The status has changed to "Needs Review" after I updated the patch. Can please you accept it?