regcache: Avoid accessing non-initialised cache
Subsystems (2)
v2 cover regcache: Avoid accessing non-initialised cache
Refactor regcache flow that populates cache on initialisation phase based on the values in HW. This makes code robust against any possible accesses to not-fully initialised caches.
Changelog v2:
- rebased on top of the latest regmap changes (Mark)
- added two more little cleanups (patches 4 & 5)
Andy Shevchenko (5):
regcache: Remove duplicate check in regcache_hw_init()
regcache: Split regcache_count_cacheable_registers() helper
regcache: Move HW readback after cache initialisation
regcache: Define iterator inside for-loop and align their types
regcache: Amend printf() specifiers when printing registers
drivers/base/regmap/internal.h | 2 +- drivers/base/regmap/regcache.c | 57 ++++++++++++++++++---------------- 2 files changed, 32 insertions(+), 27 deletions(-)
Signature · 1 line
2.50.1
PATCH v2 1/5 regcache: Remove duplicate check in regcache_hw_init()
The regcache_hw_init() is never called without preliminary check for num_reg_defaults_raw not being 0. Thus, remove duplicate in the function.
--- drivers/base/regmap/regcache.c | 3 --- 1 file changed, 3 deletions(-)
drivers/base/regmap/regcache.c +0 -3
@@ -50,9 +50,6 @@ static int regcache_hw_init(struct regmap *map)unsigned int reg, val;void *tmp_buf; -if (!map->num_reg_defaults_raw) -return -EINVAL ; -/* calculate the size of reg_defaults */ for (count =0 , i =0 ; i < map->num_reg_defaults_raw; i++) if (regmap_readable(map, i * map->reg_stride) &&
Signature · 1 line
2.50.1
PATCH v2 2/5 regcache: Split regcache_count_cacheable_registers() helper
In reply to Andy ShevchenkoThe introduced helper allows to check for the non-cacheable configurations earlier during initialisation.
--- drivers/base/regmap/regcache.c | 30 +++++++++++++++++++----------- 1 file changed, 19 insertions(+), 11 deletions(-)
drivers/base/regmap/regcache.c +19 -11
@@ -42,13 +42,10 @@ void regcache_sort_defaults(struct reg_default *defaults, unsigned int ndefaults }EXPORT_SYMBOL_GPL (regcache_sort_defaults); -static int regcache_hw_init (struct regmap *map) +static int regcache_count_cacheable_registers (struct regmap *map) { -int i, j; -int ret;int count; -unsigned int reg, val; -void *tmp_buf; +int i;/* calculate the size of reg_defaults */ for (count =0 , i =0 ; i < map->num_reg_defaults_raw; i++)
@@ -57,10 +54,18 @@ static int regcache_hw_init(struct regmap *map) count++;/* all registers are unreadable or volatile, so just bypass */ -if (!count) { +if (!count) map->cache_bypass = true; -return 0 ; - } + +return count; +} + +static int regcache_hw_init (struct regmap *map,int count) +{ +int i, j; +int ret; +unsigned int reg, val; +void *tmp_buf; map->num_reg_defaults = count; map->reg_defaults = kmalloc_objs(struct reg_default, count);
@@ -129,6 +134,7 @@ static int regcache_hw_init(struct regmap *map)int regcache_init (struct regmap *map,const struct regmap_config *config) { +int count =0 ;int ret;int i;void *tmp_buf;
@@ -193,15 +199,17 @@ int regcache_init(struct regmap *map, const struct regmap_config *config)return -ENOMEM ; map->reg_defaults = tmp_buf; }else if (map->num_reg_defaults_raw) { + count =regcache_count_cacheable_registers (map); +if (map->cache_bypass) +return 0 ; +/* Some devices such as PMICs don't have cache defaults, * we cope with this by reading back the HW registers and * crafting the cache defaults by hand. */ - ret =regcache_hw_init (map); + ret =regcache_hw_init (map, count);if (ret <0 )return ret; -if (map->cache_bypass) -return 0 ; }if (!map->max_register_is_set && map->num_reg_defaults_raw) {
Signature · 1 line
2.50.1
PATCH v2 3/5 regcache: Move HW readback after cache initialisation
In reply to Andy ShevchenkoMake sure that cache is initialised before calling any IO using regmap, this makes sure that we won't access NULL or invalid pointers in the cache which hasn't been initialised.
--- drivers/base/regmap/regcache.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-)
drivers/base/regmap/regcache.c +9 -8
@@ -202,14 +202,6 @@ int regcache_init(struct regmap *map, const struct regmap_config *config) count =regcache_count_cacheable_registers (map);if (map->cache_bypass)return 0 ; - -/* Some devices such as PMICs don't have cache defaults, -* we cope with this by reading back the HW registers and -* crafting the cache defaults by hand. -*/ - ret =regcache_hw_init (map, count); -if (ret <0 ) -return ret; }if (!map->max_register_is_set && map->num_reg_defaults_raw) {
@@ -227,6 +219,15 @@ int regcache_init(struct regmap *map, const struct regmap_config *config) goto err_free; } +/* +* Some devices such as PMICs don't have cache defaults, +* we cope with this by reading back the HW registers and +* crafting the cache defaults by hand. +*/ + ret =regcache_hw_init (map, count); +if (ret) + goto err_exit; +if (map->cache_ops->populate && (map->num_reg_defaults || map->reg_default_cb)) {dev_dbg (map->dev,"Populating %s cache\n" , map->cache_ops->name);
Signature · 1 line
2.50.1
Mark Brown
maintainerRe: [PATCH v2 3/5] regcache: Move HW readback after cache initialisation
Quoted 3 lines · Andy Shevchenko
Make sure that cache is initialised before calling any IO using regmap, this makes sure that we won't access NULL or invalid pointers in the cache which hasn't been initialised.
Either this (I suspect it's this) or the subsequent patch is introducing regressions in the ASoC tests on several of my systems in the form:
# ok 1 get_value.AT91SAMG20EK.13 # # AT91SAMG20EK.13 Inpwm8731 1-001b: ASoC error (-5): at soc_component_read_no_lock() on wm8731.1-001b for register: [0x00000004] ut Mux
https://lava.sirena.org.uk/scheduler/job/2498613#L1549
It's not absolutely everything but it's multiple systems (several with Wolfson devices, plus the Pine64 Plus which has an integrated Allwinner CODEC). I didn't check beyond seeing the failures yet.
Re: [PATCH v2 3/5] regcache: Move HW readback after cache initialisation
Quoted 12 lines · Mark Brown
Quoted 3 lines · Andy Shevchenko
Make sure that cache is initialised before calling any IO using regmap, this makes sure that we won't access NULL or invalid pointers in the cache which hasn't been initialised.
Either this (I suspect it's this) or the subsequent patch is introducing regressions in the ASoC tests on several of my systems in the form:
# ok 1 get_value.AT91SAMG20EK.13 # # AT91SAMG20EK.13 Inpwm8731 1-001b: ASoC error (-5): at soc_component_read_no_lock() on wm8731.1-001b for register: [0x00000004] ut Mux
Quoted 5 lines
https://lava.sirena.org.uk/scheduler/job/2498613#L1549
It's not absolutely everything but it's multiple systems (several with Wolfson devices, plus the Pine64 Plus which has an integrated Allwinner CODEC). I didn't check beyond seeing the failures yet.
Thanks!
I also run kunit test cases and it seems like half-failing. I'm about to check deeper what's the cause. (Obviously it works in my case on real HW, but...)
Signature · 2 lines
With Best Regards, Andy Shevchenko
Re: [PATCH v2 3/5] regcache: Move HW readback after cache initialisation
Quoted 15 lines · Andy Shevchenko
Quoted 14 lines · Mark Brown
Quoted 3 lines · Andy Shevchenko
Make sure that cache is initialised before calling any IO using regmap, this makes sure that we won't access NULL or invalid pointers in the cache which hasn't been initialised.
Either this (I suspect it's this) or the subsequent patch is introducing regressions in the ASoC tests on several of my systems in the form:
# ok 1 get_value.AT91SAMG20EK.13 # # AT91SAMG20EK.13 Inpwm8731 1-001b: ASoC error (-5): at soc_component_read_no_lock() on wm8731.1-001b for register: [0x00000004] ut Mux
Quoted 8 lines
Quoted 3 lines
It's not absolutely everything but it's multiple systems (several with Wolfson devices, plus the Pine64 Plus which has an integrated Allwinner CODEC). I didn't check beyond seeing the failures yet.
Thanks!
I also run kunit test cases and it seems like half-failing. I'm about to check deeper what's the cause. (Obviously it works in my case on real HW, but...)
At least one small amendment fixed all kunit failures for me. I will send a new version soon with also fixed patch 5.
Signature · 2 lines
With Best Regards, Andy Shevchenko
Mark Brown
maintainerRe: [PATCH v2 3/5] regcache: Move HW readback after cache initialisation
Quoted 4 lines · Andy Shevchenko
Quoted 2 lines · Andy Shevchenko
I also run kunit test cases and it seems like half-failing. I'm about to check deeper what's the cause. (Obviously it works in my case on real HW, but...)
Yeah, it worked for me on a bunch of other hardware.
Quoted 2 lines
At least one small amendment fixed all kunit failures for me. I will send a new version soon with also fixed patch 5.
Great.
PATCH v2 4/5 regcache: Define iterator inside for-loop and align their types
In reply to Andy ShevchenkoSome of the iterators may be defined inside the respective for-loop reducing the scope and potential risk of their misuse. While at it, align their types based on the type of the upper or lower limits.
--- drivers/base/regmap/internal.h | 2 +- drivers/base/regmap/regcache.c | 13 ++++++------- 2 files changed, 7 insertions(+), 8 deletions(-)
drivers/base/regmap/internal.h +1 -1
@@ -162,7 +162,7 @@ struct regmap {bool no_sync_defaults;struct reg_sequence *patch; -int patch_regs; +unsigned int patch_regs;/* if set, the regmap core can sleep */ bool can_sleep;
drivers/base/regmap/regcache.c +6 -7
@@ -44,11 +44,11 @@ EXPORT_SYMBOL_GPL(regcache_sort_defaults);static int regcache_count_cacheable_registers (struct regmap *map) { -int count; -int i; +unsigned int count;/* calculate the size of reg_defaults */ -for (count =0 , i =0 ; i < map->num_reg_defaults_raw; i++) + count =0 ; +for (unsigned int i =0 ; i < map->num_reg_defaults_raw; i++)if (regmap_readable (map, i * map->reg_stride) && !regmap_volatile (map, i * map->reg_stride)) count++;
@@ -62,7 +62,6 @@ static int regcache_count_cacheable_registers(struct regmap *map)static int regcache_hw_init (struct regmap *map,int count) { -int i, j;int ret;unsigned int reg, val;void *tmp_buf;
@@ -95,7 +94,7 @@ static int regcache_hw_init(struct regmap *map, int count) }/* fill the reg_defaults */ -for (i =0 , j =0 ; i < map->num_reg_defaults_raw; i++) { +for (unsigned int i =0 , j =0 ; i < map->num_reg_defaults_raw; i++) { reg = i * map->reg_stride;if (!regmap_readable (map, reg))
@@ -841,13 +840,13 @@ static int regcache_sync_block_raw(struct regmap *map, void *block, unsigned int block_base, unsigned int start, unsigned int end) { - unsigned int i, val; unsigned int regtmp = 0;unsigned int base =0 ;const void *data =NULL ; +unsigned int val;int ret; -for (i = start; i < end; i++) { +for (unsigned int i = start; i < end; i++) { regtmp = block_base + (i * map->reg_stride); if (!regcache_reg_present(cache_present, i) ||
Signature · 1 line
2.50.1
PATCH v2 5/5 regcache: Amend printf() specifiers when printing registers
In reply to Andy ShevchenkoIn one case the 0x is provided in the formatting string, while the rest use # for that.
In a couple of more cases a decimal signed value specifier is used.
Amend them to use %#x when register is printed. Note, for the case, when it's related to the read/write, use %x to be in align with the similar messages in regmap core.
--- drivers/base/regmap/regcache.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/base/regmap/regcache.c +3 -3
@@ -112,7 +112,7 @@ static int regcache_hw_init(struct regmap *map, int count) ret =regmap_read (map, reg, &val); map->cache_bypass = cache_bypass;if (ret !=0 ) { -dev_err (map->dev,"Failed to read %d: %d\n" , +dev_err (map->dev,"Failed to read %x: %d\n" , reg, ret); goto err_free; }
@@ -509,7 +509,7 @@ int regcache_sync_region(struct regmap *map, unsigned int min, bypass = map->cache_bypass; name = map->cache_ops->name; -dev_dbg (map->dev,"Syncing %s cache from %d-%d\n" , name, min, max); +dev_dbg (map->dev,"Syncing %s cache from %#x-%#x\n" , name, min, max);trace_regcache_sync (map, name,"start region" );
@@ -818,7 +818,7 @@ static int regcache_sync_block_raw_flush(struct regmap *map, const void **data, count = (cur - base) / map->reg_stride; -dev_dbg (map->dev,"Writing %zu bytes for %d registers from 0x%x-0x%x\n" , +dev_dbg (map->dev,"Writing %zu bytes for %d registers from %#x-%#x\n" , count * val_bytes, count, base, cur - map->reg_stride); map->cache_bypass = true;
Signature · 1 line
2.50.1
Mark Brown
maintainerRe: [PATCH v2 5/5] regcache: Amend printf() specifiers when printing registers
drivers/base/regmap/regcache.c static int regcache_sync_block_raw_flush(struct regmap *map, const void **data, · line 822
Quoted 3 lines · Andy Shevchenko
drivers/base/regmap/regcache.c +1
-1-dev_dbg (map->dev,"Writing %zu bytes for %d registers from 0x%x-0x%x\n" , +dev_dbg (map->dev,"Writing %zu bytes for %d registers from %#x-%#x\n" , count * val_bytes, count, base, cur - map->reg_stride);
I'm really not convinced this is helping legibility.
Re: [PATCH v2 5/5] regcache: Amend printf() specifiers when printing registers
Quoted 1 line · Mark Brown
On Wed, Feb 25, 2026 at 05:15:59PM +0100, Andy Shevchenko wrote:
...
Quoted 5 lines
drivers/base/regmap/regcache.c static int regcache_sync_block_raw_flush(struct regmap *map, const void **data, · line 822
Quoted 3 lines
drivers/base/regmap/regcache.c +1
-1-dev_dbg (map->dev,"Writing %zu bytes for %d registers from 0x%x-0x%x\n" , +dev_dbg (map->dev,"Writing %zu bytes for %d registers from %#x-%#x\n" , count * val_bytes, count, base, cur - map->reg_stride);I'm really not convinced this is helping legibility.
You mean the only this hunk, or the entire patch?
If the former, I have no strong opinion, can drop it as 0x%x is (almost) an equivalent to %#x. For the %d --> %x I think it needs to be applied as it makes harder to debug and follow when some of the messages use register printed in decimal.
Signature · 2 lines
With Best Regards, Andy Shevchenko
Re: [PATCH v2 5/5] regcache: Amend printf() specifiers when printing registers
Quoted 2 lines · Andy Shevchenko
Quoted 1 line · Mark Brown
On Wed, Feb 25, 2026 at 05:15:59PM +0100, Andy Shevchenko wrote:
...
Quoted 6 lines
Quoted 5 lines
drivers/base/regmap/regcache.c static int regcache_sync_block_raw_flush(struct regmap *map, const void **data, · line 822
Quoted 3 lines
drivers/base/regmap/regcache.c +1
-1-dev_dbg (map->dev,"Writing %zu bytes for %d registers from 0x%x-0x%x\n" , +dev_dbg (map->dev,"Writing %zu bytes for %d registers from %#x-%#x\n" , count * val_bytes, count, base, cur - map->reg_stride);I'm really not convinced this is helping legibility.
Quoted 6 lines
You mean the only this hunk, or the entire patch?
If the former, I have no strong opinion, can drop it as 0x%x is (almost) an equivalent to %#x. For the %d --> %x I think it needs to be applied as
it makes harder to debug and follow when some of the messages use register
Under 'it' I meant here the decimal printing versus hexadecimal.
Quoted 1 line
printed in decimal.
Signature · 2 lines
With Best Regards, Andy Shevchenko
Mark Brown
maintainerRe: [PATCH v2 5/5] regcache: Amend printf() specifiers when printing registers
In reply to Andy ShevchenkoQuoted 10 lines · Andy Shevchenko
Quoted 7 lines · Mark Brown
drivers/base/regmap/regcache.c static int regcache_sync_block_raw_flush(struct regmap *map, const void **data, · line 822
Quoted 3 lines · Andy Shevchenko
drivers/base/regmap/regcache.c +1
-1-dev_dbg (map->dev,"Writing %zu bytes for %d registers from 0x%x-0x%x\n" , +dev_dbg (map->dev,"Writing %zu bytes for %d registers from %#x-%#x\n" , count * val_bytes, count, base, cur - map->reg_stride);I'm really not convinced this is helping legibility.
You mean the only this hunk, or the entire patch?
Just that hunk, it's new and unusual, reads less cleanly and I can't identify any problem it solves.
Mark Brown
maintainerRe: (subset) [PATCH v2 0/5] regcache: Avoid accessing non-initialised cache
In reply to Andy ShevchenkoQuoted 9 lines · Andy Shevchenko
Refactor regcache flow that populates cache on initialisation phase based on the values in HW. This makes code robust against any possible accesses to not-fully initialised caches.
Changelog v2:
- rebased on top of the latest regmap changes (Mark)
- added two more little cleanups (patches 4 & 5)[...]
Applied to
https://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap.git for-next
Thanks!
[1/5] regcache: Remove duplicate check in regcache_hw_init()
commit: 1fcf171178f021ef2005d47e281544624b93c5a5
[2/5] regcache: Split regcache_count_cacheable_registers() helper
commit: c2bcf62ca75c541ec4297e6ff02a68ddc2e02029[3/5] regcache: Move HW readback after cache initialisation
(no commit info)
[4/5] regcache: Define iterator inside for-loop and align their types
(no commit info)
All being well this means that it will be integrated into the linux-next tree (usually sometime in the next 24 hours) and sent to Linus during the next merge window (or sooner if it is a bug fix), however if problems are discovered then the patch may be dropped or reverted.
You may get further e-mails resulting from automated or manual testing and review of the tree, please engage with people reporting problems and send followup patches addressing any issues that are reported if needed.
If any updates are required or you are submitting further changes they should be sent as incremental updates against current git, existing patches will not be replaced.
Please add any relevant lists and maintainers to the CCs when replying to this mail.
Thanks,
Mark