Here's a small series that lets us enable COMPILE_TEST for the pic32 clk
driver so that this driver can be compiled on all architectures in the
future to help avoid breakage.

Signed-off-by: Brian Masney <bmasney@redhat.com>
---
Brian Masney (3):
      clk: microchip: core: update to use div64_ul() instead of do_div()
      clk: microchip: core: change asm nop calls to nop()
      clk: microchip: core: allow driver to be compiled with COMPILE_TEST

 drivers/clk/microchip/Kconfig    |  2 +-
 drivers/clk/microchip/clk-core.c | 13 +++++++------
 2 files changed, 8 insertions(+), 7 deletions(-)
---
base-commit: 6de23f81a5e08be8fbf5e8d7e9febc72a5b5f27f
change-id: 20260222-clk-microchip-pic32-0e5b00f122a0

Best regards,
-- 
Brian Masney <bmasney@redhat.com>

On Sun, Feb 22, 2026 at 06:51:03PM -0500, Brian Masney wrote:
> Here's a small series that lets us enable COMPILE_TEST for the pic32 clk
> driver so that this driver can be compiled on all architectures in the
> future to help avoid breakage.

Why do you care about this platform? Originally I thought you cared cos
you were getting involved in clock subsystem stuff, but I see you doing
stuff with it in pinctrl etc. Is that just general cleanup work, or do
you actually have an interest in the platform?

Also, seeing "clk: microchip: core" in my inbox triggers me, cos the
stuff I work on uses "microchip: core" a lot for FPGA IP cores. Can we
do:

-- >8 --
From 5e972ac119c87257ce141382019f2b58ebd53359 Mon Sep 17 00:00:00 2001
From: Conor Dooley <conor.dooley@microchip.com>
Date: Tue, 24 Feb 2026 16:59:53 +0000
Subject: [PATCH] clk: microchip: rename clk-core to clk-pic32

clk-core is a confusingly generic name, since it is only used by a
single platform and it uses very similar naming to the "soft" IP cores
for use in FPGA fabric (CoreClock or similar is what that would be
called, although nothing like that exists right now) that the FPGA
business unit produces. Rename it to clk-pic32, matching the prefix
used by most functions in the driver. As far as I can tell, impact
on whatever users may (or may not...) exist for the platform is minimal
as it's built-in only and the functions are called directly from
clk-pic32mzda.c

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
 drivers/clk/microchip/Makefile                    | 2 +-
 drivers/clk/microchip/{clk-core.c => clk-pic32.c} | 2 +-
 drivers/clk/microchip/{clk-core.h => clk-pic32.h} | 0
 drivers/clk/microchip/clk-pic32mzda.c             | 2 +-
 4 files changed, 3 insertions(+), 3 deletions(-)
 rename drivers/clk/microchip/{clk-core.c => clk-pic32.c} (99%)
 rename drivers/clk/microchip/{clk-core.h => clk-pic32.h} (100%)

diff --git a/drivers/clk/microchip/Makefile b/drivers/clk/microchip/Makefile
index 13250e04e46cf..8e60bc1a03ae7 100644
--- a/drivers/clk/microchip/Makefile
+++ b/drivers/clk/microchip/Makefile
@@ -1,5 +1,5 @@
 # SPDX-License-Identifier: GPL-2.0-only
-obj-$(CONFIG_COMMON_CLK_PIC32) += clk-core.o
+obj-$(CONFIG_COMMON_CLK_PIC32) += clk-pic32.o
 obj-$(CONFIG_PIC32MZDA) += clk-pic32mzda.o
 obj-$(CONFIG_MCHP_CLK_MPFS) += clk-mpfs.o
 obj-$(CONFIG_MCHP_CLK_MPFS) += clk-mpfs-ccc.o
diff --git a/drivers/clk/microchip/clk-core.c b/drivers/clk/microchip/clk-pic32.c
similarity index 99%
rename from drivers/clk/microchip/clk-core.c
rename to drivers/clk/microchip/clk-pic32.c
index 692152b5094e0..9d128fba2cdea 100644
--- a/drivers/clk/microchip/clk-core.c
+++ b/drivers/clk/microchip/clk-pic32.c
@@ -11,7 +11,7 @@
 #include <linux/iopoll.h>
 #include <linux/platform_data/pic32.h>
 
-#include "clk-core.h"
+#include "clk-pic32.h"
 
 /* OSCCON Reg fields */
 #define OSC_CUR_MASK		0x07
diff --git a/drivers/clk/microchip/clk-core.h b/drivers/clk/microchip/clk-pic32.h
similarity index 100%
rename from drivers/clk/microchip/clk-core.h
rename to drivers/clk/microchip/clk-pic32.h
diff --git a/drivers/clk/microchip/clk-pic32mzda.c b/drivers/clk/microchip/clk-pic32mzda.c
index 27599829ea40a..e11cbdd982a61 100644
--- a/drivers/clk/microchip/clk-pic32mzda.c
+++ b/drivers/clk/microchip/clk-pic32mzda.c
@@ -14,7 +14,7 @@
 #include <linux/platform_device.h>
 #include <asm/traps.h>
 
-#include "clk-core.h"
+#include "clk-pic32.h"
 
 /* FRC Postscaler */
 #define OSC_FRCDIV_MASK		0x07
-- 
2.51.0

On Tue, Feb 24, 2026 at 05:10:42PM +0000, Conor Dooley wrote:
> On Sun, Feb 22, 2026 at 06:51:03PM -0500, Brian Masney wrote:
> > Here's a small series that lets us enable COMPILE_TEST for the pic32 clk
> > driver so that this driver can be compiled on all architectures in the
> > future to help avoid breakage.
> 
> Why do you care about this platform? Originally I thought you cared cos
> you were getting involved in clock subsystem stuff, but I see you doing
> stuff with it in pinctrl etc. Is that just general cleanup work, or do
> you actually have an interest in the platform?

My sole interest is to clean up the clk driver, which spilled into
larger work to clean the include header to allow this driver to be
compiled on other architectures. If I was going to do this for the
clk driver, it wasn't that much work to do the other pic32 driver
cleanups while I was there.

Honestly, I hope to done with all of these pic32 drivers after what I
posted this week has been merged.

> Also, seeing "clk: microchip: core" in my inbox triggers me, cos the
> stuff I work on uses "microchip: core" a lot for FPGA IP cores. Can we
> do:
> 
> -- >8 --
> From 5e972ac119c87257ce141382019f2b58ebd53359 Mon Sep 17 00:00:00 2001
> From: Conor Dooley <conor.dooley@microchip.com>
> Date: Tue, 24 Feb 2026 16:59:53 +0000
> Subject: [PATCH] clk: microchip: rename clk-core to clk-pic32
> 
> clk-core is a confusingly generic name, since it is only used by a
> single platform and it uses very similar naming to the "soft" IP cores
> for use in FPGA fabric (CoreClock or similar is what that would be
> called, although nothing like that exists right now) that the FPGA
> business unit produces. Rename it to clk-pic32, matching the prefix
> used by most functions in the driver. As far as I can tell, impact
> on whatever users may (or may not...) exist for the platform is minimal
> as it's built-in only and the functions are called directly from
> clk-pic32mzda.c
> 
> Signed-off-by: Conor Dooley <conor.dooley@microchip.com>

That sounds good to me. Please post that as a separate patch.

Brian

On Tue, Feb 24, 2026 at 12:27:27PM -0500, Brian Masney wrote:
> On Tue, Feb 24, 2026 at 05:10:42PM +0000, Conor Dooley wrote:
> > On Sun, Feb 22, 2026 at 06:51:03PM -0500, Brian Masney wrote:
> > > Here's a small series that lets us enable COMPILE_TEST for the pic32 clk
> > > driver so that this driver can be compiled on all architectures in the
> > > future to help avoid breakage.
> > 
> > Why do you care about this platform? Originally I thought you cared cos
> > you were getting involved in clock subsystem stuff, but I see you doing
> > stuff with it in pinctrl etc. Is that just general cleanup work, or do
> > you actually have an interest in the platform?
> 
> My sole interest is to clean up the clk driver, which spilled into
> larger work to clean the include header to allow this driver to be
> compiled on other architectures. If I was going to do this for the
> clk driver, it wasn't that much work to do the other pic32 driver
> cleanups while I was there.
> 
> Honestly, I hope to done with all of these pic32 drivers after what I
> posted this week has been merged.

Not to pooh pooh your efforts, but I asking as I am wondering whether
the whole platform should be deleted. I'm gonna ask internally about it,
but none of the authors are around anymore and seeing
| git log --grep "pic32" --author="@microchip.com" --pretty="%h%x09%an%x09%ad%x09%s" -n 2 --date=short
| 5e972ac119c87   Conor Dooley    2026-02-24      clk: microchip: rename clk-core to clk-pic32
| a726f1d2dd4fe   Purna Chandra Mandal    2016-06-02      MIPS: pic32mzda: Fix linker error for pic32_get_pbclk()
is bad omen.

On Tue, Feb 24, 2026 at 05:36:25PM +0000, Conor Dooley wrote:
> On Tue, Feb 24, 2026 at 12:27:27PM -0500, Brian Masney wrote:
> > On Tue, Feb 24, 2026 at 05:10:42PM +0000, Conor Dooley wrote:
> > > On Sun, Feb 22, 2026 at 06:51:03PM -0500, Brian Masney wrote:
> > > > Here's a small series that lets us enable COMPILE_TEST for the pic32 clk
> > > > driver so that this driver can be compiled on all architectures in the
> > > > future to help avoid breakage.
> > > 
> > > Why do you care about this platform? Originally I thought you cared cos
> > > you were getting involved in clock subsystem stuff, but I see you doing
> > > stuff with it in pinctrl etc. Is that just general cleanup work, or do
> > > you actually have an interest in the platform?
> > 
> > My sole interest is to clean up the clk driver, which spilled into
> > larger work to clean the include header to allow this driver to be
> > compiled on other architectures. If I was going to do this for the
> > clk driver, it wasn't that much work to do the other pic32 driver
> > cleanups while I was there.
> > 
> > Honestly, I hope to done with all of these pic32 drivers after what I
> > posted this week has been merged.
> 
> Not to pooh pooh your efforts, but I asking as I am wondering whether
> the whole platform should be deleted. I'm gonna ask internally about it,
> but none of the authors are around anymore and seeing
> | git log --grep "pic32" --author="@microchip.com" --pretty="%h%x09%an%x09%ad%x09%s" -n 2 --date=short
> | 5e972ac119c87   Conor Dooley    2026-02-24      clk: microchip: rename clk-core to clk-pic32
> | a726f1d2dd4fe   Purna Chandra Mandal    2016-06-02      MIPS: pic32mzda: Fix linker error for pic32_get_pbclk()
> is bad omen.

Yea, that's not a good sign. No objections from me.

Brian