pinctrl-single: fix pcs_parse_pinconf() return value
Subsystems (3)
PATCH v2 1/1 pinctrl-single: fix pcs_parse_pinconf() return value
This patch causes pcs_parse_pinconf() to return -ENOTSUPP when no pinctrl_map is added. The current behavior is to return 0 when !PCS_HAS_PINCONF or !nconfs. Thus pcs_parse_one_pinctrl_entry() incorrectly assumes that a map was added and sets num_maps = 2.
Analysis:
=========
The function pcs_parse_one_pinctrl_entry() calls pcs_parse_pinconf() if PCS_HAS_PINCONF is enabled. The function pcs_parse_pinconf() returns 0 to indicate there was no error and num_maps is then set to 2:
980 static int pcs_parse_one_pinctrl_entry(struct pcs_device *pcs,
981 struct device_node *np,
982 struct pinctrl_map **map,
983 unsigned *num_maps,
984 const char **pgnames)
985 {
<snip>
1053 (*map)->type = PIN_MAP_TYPE_MUX_GROUP;
1054 (*map)->data.mux.group = np->name;
1055 (*map)->data.mux.function = np->name;1056
1057 if (PCS_HAS_PINCONF && function) {
1058 res = pcs_parse_pinconf(pcs, np, function, map);
1059 if (res)
1060 goto free_pingroups;
1061 *num_maps = 2;
1062 } else {
1063 *num_maps = 1;
1064 }
However, pcs_parse_pinconf() will also return 0 if !PCS_HAS_PINCONF or !nconfs. I believe these conditions should indicate that no map was added by returning -ENOTSUPP. Otherwise pcs_parse_one_pinctrl_entry() will set num_maps = 2 even though no maps were successfully added, as it does not reach "m++" on line 940:
895 static int pcs_parse_pinconf(struct pcs_device *pcs, struct device_node *np,
896 struct pcs_function *func,
897 struct pinctrl_map **map)
898
899 {
900 struct pinctrl_map *m = *map;
<snip>
917 /* If pinconf isn't supported, don't parse properties in below. */
918 if (!PCS_HAS_PINCONF)
919 return 0;
920
921 /* cacluate how much properties are supported in current node */
922 for (i = 0; i < ARRAY_SIZE(prop2); i++) {
923 if (of_find_property(np, prop2[i].name, NULL))
924 nconfs++;
925 }
926 for (i = 0; i < ARRAY_SIZE(prop4); i++) {
927 if (of_find_property(np, prop4[i].name, NULL))
928 nconfs++;
929 }
930 if (!nconfs)
919 return 0;
932
933 func->conf = devm_kcalloc(pcs->dev,
934 nconfs, sizeof(struct pcs_conf_vals),
935 GFP_KERNEL);
936 if (!func->conf)
937 return -ENOMEM;
938 func->nconfs = nconfs; 939 conf = &(func->conf[0]); 940 m++;
This situtation will cause a boot failure [0] on the BeagleBone Black (AM3358) when am33xx_pinmux node in arch/arm/boot/dts/am33xx-l4.dtsi has compatible = "pinconf-single" instead of "pinctrl-single".
The patch fixes this issue by returning -ENOSUPP when !PCS_HAS_PINCONF or !nconfs, so that pcs_parse_one_pinctrl_entry() will know that no map was added.
Logic is also added to pcs_parse_one_pinctrl_entry() to distinguish between -ENOSUPP and other errors. In the case of -ENOSUPP, num_maps is set to 1 as it is valid for pinconf to be enabled and a given pin group to not any pinconf properties.
[0] https://lore.kernel.org/linux-omap/20200529175544.GA3766151@x1/
---
changes from V1 [0]:
- if pcs_parse_pinconf() returns -ENOSUPP, then set num_maps to 1 and proceed normally as it is valid for group to have no pinconf props
- added Fixes: tag thanks to Gustavo A. R. Silva
[0] https://lore.kernel.org/linux-omap/20200531204147.GA664833@x1/
drivers/pinctrl/pinctrl-single.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-)
drivers/pinctrl/pinctrl-single.c +7 -4
@@ -916,7 +916,7 @@ static int pcs_parse_pinconf(struct pcs_device *pcs, struct device_node *np,/* If pinconf isn't supported, don't parse properties in below. */ if (!PCS_HAS_PINCONF ) -return 0 ; +return -ENOTSUPP ;/* cacluate how much properties are supported in current node */ for (i =0 ; i <ARRAY_SIZE (prop2); i++) {
@@ -928,7 +928,7 @@ static int pcs_parse_pinconf(struct pcs_device *pcs, struct device_node *np, nconfs++; }if (!nconfs) -return 0 ; +return -ENOTSUPP ; func->conf = devm_kcalloc(pcs->dev, nconfs, sizeof(struct pcs_conf_vals),
@@ -1056,9 +1056,12 @@ static int pcs_parse_one_pinctrl_entry(struct pcs_device *pcs,if (PCS_HAS_PINCONF && function) { res =pcs_parse_pinconf (pcs, np, function, map); -if (res) +if (res ==0 ) + *num_maps =2 ; +else if (res == -ENOTSUPP ) + *num_maps =1 ; +else goto free_pingroups; - *num_maps =2 ; }else { *num_maps =1 ; }
Signature · 1 line
2.25.1
Quoted 4 lines · Drew Fustini
This patch causes
pcs_parse_pinconf()to return-ENOTSUPPwhen nopinctrl_mapis added. The current behavior is to return 0 when !PCS_HAS_PINCONF or !nconfs. Thuspcs_parse_one_pinctrl_entry()incorrectly assumes that a map was added and setsnum_maps= 2.
...
Quoted 1 line
Fixes 9dddb4df90d1 ("pinctrl: single: support generic pinconf")
It would be good to get an ack/tested-by from Haojian for this.
The patch looks right to me:
Quoted 82 lines · Drew Fustini
This patch causes
pcs_parse_pinconf()to return-ENOTSUPPwhen nopinctrl_mapis added. The current behavior is to return 0 when !PCS_HAS_PINCONF or !nconfs. Thuspcs_parse_one_pinctrl_entry()incorrectly assumes that a map was added and setsnum_maps= 2.Analysis:
=========
The functionpcs_parse_one_pinctrl_entry()callspcs_parse_pinconf()ifPCS_HAS_PINCONFis enabled. The functionpcs_parse_pinconf()returns 0 to indicate there was no error andnum_mapsis then set to 2:980 static int pcs_parse_one_pinctrl_entry(struct
pcs_device*pcs,981 struct device_node *np, 982 struct pinctrl_map **map, 983 unsigned *num_maps, 984 const char **pgnames) 985 { <snip> 1053 (*map)->type = PIN_MAP_TYPE_MUX_GROUP; 1054 (*map)->data.mux.group = np->name; 1055 (*map)->data.mux.function = np->name;1056
1057 if (PCS_HAS_PINCONF&& function) {
1058 res =pcs_parse_pinconf(pcs, np, function, map);
1059 if (res)
1060 gotofree_pingroups;
1061*num_maps= 2;
1062 } else {
1063*num_maps= 1;
1064 }However,
pcs_parse_pinconf()will also return 0 if !PCS_HAS_PINCONF or !nconfs. I believe these conditions should indicate that no map was added by returning-ENOTSUPP. Otherwisepcs_parse_one_pinctrl_entry()will setnum_maps= 2 even though no maps were successfully added, as it does not reach "m++" on line 940:895 static int pcs_parse_pinconf(struct
pcs_device*pcs, structdevice_node*np,
896 structpcs_function*func,
897 structpinctrl_map**map)
898
899 {
900 structpinctrl_map*m = *map;
<snip>
917 /* If pinconf isn't supported, don't parse properties in below. */
918 if (!PCS_HAS_PINCONF)
919 return 0;
920
921 /* cacluate how much properties are supported in current node */
922 for (i = 0; i <ARRAY_SIZE(prop2); i++) {
923 if (of_find_property(np, prop2[i].name, NULL))
924 nconfs++;
925 }
926 for (i = 0; i <ARRAY_SIZE(prop4); i++) {
927 if (of_find_property(np, prop4[i].name, NULL))
928 nconfs++;
929 }
930 if (!nconfs)
919 return 0;
932
933func->conf= devm_kcalloc(pcs->dev,
934 nconfs,sizeof(struct pcs_conf_vals),
935GFP_KERNEL);
936 if (!func->conf)
937 return-ENOMEM;938 func->nconfs = nconfs; 939 conf = &(func->conf[0]); 940 m++;This situtation will cause a boot failure [0] on the BeagleBone Black (AM3358) when
am33xx_pinmuxnode inarch/arm/boot/dts/am33xx-l4.dtsihas compatible = "pinconf-single" instead of "pinctrl-single".The patch fixes this issue by returning
-ENOSUPPwhen !PCS_HAS_PINCONF or !nconfs, so thatpcs_parse_one_pinctrl_entry()will know that no map was added.Logic is also added to
pcs_parse_one_pinctrl_entry()to distinguish between-ENOSUPPand other errors. In the case of-ENOSUPP,num_mapsis set to 1 as it is valid for pinconf to be enabled and a given pin group to not any pinconf properties.
Quoted 4 lines
[0] https://lore.kernel.org/linux-omap/20200529175544.GA3766151@x1/
Fixes 9dddb4df90d1 ("pinctrl: single: support generic pinconf")Signed-off-by Drew Fustini <[email protected]>
Patch applied as non-critical (for-next) fix.
If there is no hurry let's merge it this way with lots of testing along the way.
Yours,
Linus Walleij
Quoted 89 lines · Linus Walleij
Quoted 82 lines · Drew Fustini
This patch causes
pcs_parse_pinconf()to return-ENOTSUPPwhen nopinctrl_mapis added. The current behavior is to return 0 when !PCS_HAS_PINCONF or !nconfs. Thuspcs_parse_one_pinctrl_entry()incorrectly assumes that a map was added and setsnum_maps= 2.Analysis:
=========
The functionpcs_parse_one_pinctrl_entry()callspcs_parse_pinconf()ifPCS_HAS_PINCONFis enabled. The functionpcs_parse_pinconf()returns 0 to indicate there was no error andnum_mapsis then set to 2:980 static int pcs_parse_one_pinctrl_entry(struct
pcs_device*pcs,981 struct device_node *np, 982 struct pinctrl_map **map, 983 unsigned *num_maps, 984 const char **pgnames) 985 { <snip> 1053 (*map)->type = PIN_MAP_TYPE_MUX_GROUP; 1054 (*map)->data.mux.group = np->name; 1055 (*map)->data.mux.function = np->name;1056
1057 if (PCS_HAS_PINCONF&& function) {
1058 res =pcs_parse_pinconf(pcs, np, function, map);
1059 if (res)
1060 gotofree_pingroups;
1061*num_maps= 2;
1062 } else {
1063*num_maps= 1;
1064 }However,
pcs_parse_pinconf()will also return 0 if !PCS_HAS_PINCONF or !nconfs. I believe these conditions should indicate that no map was added by returning-ENOTSUPP. Otherwisepcs_parse_one_pinctrl_entry()will setnum_maps= 2 even though no maps were successfully added, as it does not reach "m++" on line 940:895 static int pcs_parse_pinconf(struct
pcs_device*pcs, structdevice_node*np,
896 structpcs_function*func,
897 structpinctrl_map**map)
898
899 {
900 structpinctrl_map*m = *map;
<snip>
917 /* If pinconf isn't supported, don't parse properties in below. */
918 if (!PCS_HAS_PINCONF)
919 return 0;
920
921 /* cacluate how much properties are supported in current node */
922 for (i = 0; i <ARRAY_SIZE(prop2); i++) {
923 if (of_find_property(np, prop2[i].name, NULL))
924 nconfs++;
925 }
926 for (i = 0; i <ARRAY_SIZE(prop4); i++) {
927 if (of_find_property(np, prop4[i].name, NULL))
928 nconfs++;
929 }
930 if (!nconfs)
919 return 0;
932
933func->conf= devm_kcalloc(pcs->dev,
934 nconfs,sizeof(struct pcs_conf_vals),
935GFP_KERNEL);
936 if (!func->conf)
937 return-ENOMEM;938 func->nconfs = nconfs; 939 conf = &(func->conf[0]); 940 m++;This situtation will cause a boot failure [0] on the BeagleBone Black (AM3358) when
am33xx_pinmuxnode inarch/arm/boot/dts/am33xx-l4.dtsihas compatible = "pinconf-single" instead of "pinctrl-single".The patch fixes this issue by returning
-ENOSUPPwhen !PCS_HAS_PINCONF or !nconfs, so thatpcs_parse_one_pinctrl_entry()will know that no map was added.Logic is also added to
pcs_parse_one_pinctrl_entry()to distinguish between-ENOSUPPand other errors. In the case of-ENOSUPP,num_mapsis set to 1 as it is valid for pinconf to be enabled and a given pin group to not any pinconf properties.Quoted 4 lines
[0] https://lore.kernel.org/linux-omap/20200529175544.GA3766151@x1/
Fixes 9dddb4df90d1 ("pinctrl: single: support generic pinconf")Signed-off-by Drew Fustini <[email protected]>
Quoted 7 lines
Patch applied as non-critical (for-next) fix.
If there is no hurry let's merge it this way with lots of testing along the way.
Yours,
Linus Walleij
Thanks, I agree more testing is a good idea.
In particular, do you have a way to followup with Haojian Zhuang within Linaro?
Haojian added the generic pinconf support back in 2013, so it would be good to get his review.
Also, neither Tony or I have the Hikey hardware to test.
The most important to test would be those which include a .dtsi with "pinconf-single" compatible. I do plan to add pinconf-single to the AM3358 BeagleBone (which is what I am testing on), but the current mainline users are:
arch/arm/boot/dts/hi3620.dtsi
|- arch/arm/boot/dts/hi3620-hi4511.dts
arch/arm/boot/dts/pxa3xx.dtsi
|- arch/arm/boot/dts/pxa300-raumfeld-common.dtsi
|- arch/arm/boot/dts/pxa300-raumfeld-connector.dts
arch/arm/boot/dts/pxa300-raumfeld-controller.dts arch/arm/boot/dts/pxa300-raumfeld-speaker-l.dts arch/arm/boot/dts/pxa300-raumfeld-speaker-m.dts arch/arm/boot/dts/pxa300-raumfeld-speaker-one.dts arch/arm/boot/dts/pxa300-raumfeld-speaker-s.dts
I am cc'ing Daniel Mack and Robert Jarzmi as they are listed as maintainers for the pxa300 related dts files.
Those platforms using compatible = "pinctrl-single" should not be effected by this patch:
arch/arm/boot/dts/am33xx-l4.dtsi (I have changed to pinconf for test) arch/arm/boot/dts/da850.dtsiarch/arm/boot/dts/dm814x.dtsi arch/arm/boot/dts/dm816x.dtsi arch/arm/boot/dts/hi3620.dtsi arch/arm/boot/dts/keystone-k2g.dtsi arch/arm/boot/dts/keystone-k2l.dtsi arch/arm/boot/dts/omap3-gta04.dtsi
Thanks,
Drew
Quoted 95 lines · Drew Fustini
Quoted 94 lines · Linus Walleij
Quoted 82 lines · Drew Fustini
This patch causes
pcs_parse_pinconf()to return-ENOTSUPPwhen nopinctrl_mapis added. The current behavior is to return 0 when !PCS_HAS_PINCONF or !nconfs. Thuspcs_parse_one_pinctrl_entry()incorrectly assumes that a map was added and setsnum_maps= 2.Analysis:
=========
The functionpcs_parse_one_pinctrl_entry()callspcs_parse_pinconf()ifPCS_HAS_PINCONFis enabled. The functionpcs_parse_pinconf()returns 0 to indicate there was no error andnum_mapsis then set to 2:980 static int pcs_parse_one_pinctrl_entry(struct
pcs_device*pcs,981 struct device_node *np, 982 struct pinctrl_map **map, 983 unsigned *num_maps, 984 const char **pgnames) 985 { <snip> 1053 (*map)->type = PIN_MAP_TYPE_MUX_GROUP; 1054 (*map)->data.mux.group = np->name; 1055 (*map)->data.mux.function = np->name;1056
1057 if (PCS_HAS_PINCONF&& function) {
1058 res =pcs_parse_pinconf(pcs, np, function, map);
1059 if (res)
1060 gotofree_pingroups;
1061*num_maps= 2;
1062 } else {
1063*num_maps= 1;
1064 }However,
pcs_parse_pinconf()will also return 0 if !PCS_HAS_PINCONF or !nconfs. I believe these conditions should indicate that no map was added by returning-ENOTSUPP. Otherwisepcs_parse_one_pinctrl_entry()will setnum_maps= 2 even though no maps were successfully added, as it does not reach "m++" on line 940:895 static int pcs_parse_pinconf(struct
pcs_device*pcs, structdevice_node*np,
896 structpcs_function*func,
897 structpinctrl_map**map)
898
899 {
900 structpinctrl_map*m = *map;
<snip>
917 /* If pinconf isn't supported, don't parse properties in below. */
918 if (!PCS_HAS_PINCONF)
919 return 0;
920
921 /* cacluate how much properties are supported in current node */
922 for (i = 0; i <ARRAY_SIZE(prop2); i++) {
923 if (of_find_property(np, prop2[i].name, NULL))
924 nconfs++;
925 }
926 for (i = 0; i <ARRAY_SIZE(prop4); i++) {
927 if (of_find_property(np, prop4[i].name, NULL))
928 nconfs++;
929 }
930 if (!nconfs)
919 return 0;
932
933func->conf= devm_kcalloc(pcs->dev,
934 nconfs,sizeof(struct pcs_conf_vals),
935GFP_KERNEL);
936 if (!func->conf)
937 return-ENOMEM;938 func->nconfs = nconfs; 939 conf = &(func->conf[0]); 940 m++;This situtation will cause a boot failure [0] on the BeagleBone Black (AM3358) when
am33xx_pinmuxnode inarch/arm/boot/dts/am33xx-l4.dtsihas compatible = "pinconf-single" instead of "pinctrl-single".The patch fixes this issue by returning
-ENOSUPPwhen !PCS_HAS_PINCONF or !nconfs, so thatpcs_parse_one_pinctrl_entry()will know that no map was added.Logic is also added to
pcs_parse_one_pinctrl_entry()to distinguish between-ENOSUPPand other errors. In the case of-ENOSUPP,num_mapsis set to 1 as it is valid for pinconf to be enabled and a given pin group to not any pinconf properties.Quoted 4 lines
[0] https://lore.kernel.org/linux-omap/20200529175544.GA3766151@x1/
Fixes 9dddb4df90d1 ("pinctrl: single: support generic pinconf")Signed-off-by Drew Fustini <[email protected]>Patch applied as non-critical (for-next) fix.
If there is no hurry let's merge it this way with lots of testing along the way.
Quoted 8 lines
Quoted 2 lines
Yours,
Linus WalleijThanks, I agree more testing is a good idea.
In particular, do you have a way to followup with Haojian Zhuang within Linaro?
Linus - do you have a way to contact Haojian Zhuan?
I found them on Freenode but they have been idle since beginning of June. Last email I can find on the mailing list is from March.
Tony and I would both like someone familiar with Hisilicon to comment on this patch as it is using pinctrl-single.
Thanks,
Drew
Quoted 136 lines · Drew Fustini
This patch causes
pcs_parse_pinconf()to return-ENOTSUPPwhen nopinctrl_mapis added. The current behavior is to return 0 when !PCS_HAS_PINCONF or !nconfs. Thuspcs_parse_one_pinctrl_entry()incorrectly assumes that a map was added and setsnum_maps= 2.Analysis:
=========
The functionpcs_parse_one_pinctrl_entry()callspcs_parse_pinconf()ifPCS_HAS_PINCONFis enabled. The functionpcs_parse_pinconf()returns 0 to indicate there was no error andnum_mapsis then set to 2:980 static int pcs_parse_one_pinctrl_entry(struct
pcs_device*pcs,981 struct device_node *np, 982 struct pinctrl_map **map, 983 unsigned *num_maps, 984 const char **pgnames) 985 { <snip> 1053 (*map)->type = PIN_MAP_TYPE_MUX_GROUP; 1054 (*map)->data.mux.group = np->name; 1055 (*map)->data.mux.function = np->name;1056
1057 if (PCS_HAS_PINCONF&& function) {
1058 res =pcs_parse_pinconf(pcs, np, function, map);
1059 if (res)
1060 gotofree_pingroups;
1061*num_maps= 2;
1062 } else {
1063*num_maps= 1;
1064 }However,
pcs_parse_pinconf()will also return 0 if !PCS_HAS_PINCONF or !nconfs. I believe these conditions should indicate that no map was added by returning-ENOTSUPP. Otherwisepcs_parse_one_pinctrl_entry()will setnum_maps= 2 even though no maps were successfully added, as it does not reach "m++" on line 940:895 static int pcs_parse_pinconf(struct
pcs_device*pcs, structdevice_node*np,
896 structpcs_function*func,
897 structpinctrl_map**map)
898
899 {
900 structpinctrl_map*m = *map;
<snip>
917 /* If pinconf isn't supported, don't parse properties in below. */
918 if (!PCS_HAS_PINCONF)
919 return 0;
920
921 /* cacluate how much properties are supported in current node */
922 for (i = 0; i <ARRAY_SIZE(prop2); i++) {
923 if (of_find_property(np, prop2[i].name, NULL))
924 nconfs++;
925 }
926 for (i = 0; i <ARRAY_SIZE(prop4); i++) {
927 if (of_find_property(np, prop4[i].name, NULL))
928 nconfs++;
929 }
930 if (!nconfs)
919 return 0;
932
933func->conf= devm_kcalloc(pcs->dev,
934 nconfs,sizeof(struct pcs_conf_vals),
935GFP_KERNEL);
936 if (!func->conf)
937 return-ENOMEM;938 func->nconfs = nconfs; 939 conf = &(func->conf[0]); 940 m++;This situtation will cause a boot failure [0] on the BeagleBone Black (AM3358) when
am33xx_pinmuxnode inarch/arm/boot/dts/am33xx-l4.dtsihas compatible = "pinconf-single" instead of "pinctrl-single".The patch fixes this issue by returning
-ENOSUPPwhen !PCS_HAS_PINCONF or !nconfs, so thatpcs_parse_one_pinctrl_entry()will know that no map was added.Logic is also added to
pcs_parse_one_pinctrl_entry()to distinguish between-ENOSUPPand other errors. In the case of-ENOSUPP,num_mapsis set to 1 as it is valid for pinconf to be enabled and a given pin group to not any pinconf properties.[0] https://lore.kernel.org/linux-omap/20200529175544.GA3766151@x1/
Fixes 9dddb4df90d1 ("pinctrl: single: support generic pinconf")Signed-off-by Drew Fustini <[email protected]>---
changes from V1 [0]:
- ifpcs_parse_pinconf()returns-ENOSUPP, then setnum_mapsto 1 and proceed normally as it is valid for group to have no pinconf props
- added Fixes: tag thanks to Gustavo A. R. Silva[0] https://lore.kernel.org/linux-omap/20200531204147.GA664833@x1/
drivers/pinctrl/pinctrl-single.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-)drivers/pinctrl/pinctrl-single.c +7
-5@@ -916,7 +916,7 @@ static int pcs_parse_pinconf(struct pcs_device *pcs, struct device_node *np,/* If pinconf isn't supported, don't parse properties in below. */ if (!PCS_HAS_PINCONF ) -return 0 ; +return -ENOTSUPP ;/* cacluate how much properties are supported in current node */ for (i =0 ; i <ARRAY_SIZE (prop2); i++) {@@ -928,7 +928,7 @@ static int pcs_parse_pinconf(struct pcs_device *pcs, struct device_node *np, nconfs++; }if (!nconfs) -return 0 ; +return -ENOTSUPP ; func->conf = devm_kcalloc(pcs->dev, nconfs, sizeof(struct pcs_conf_vals),@@ -1056,9 +1056,12 @@ static int pcs_parse_one_pinctrl_entry(struct pcs_device *pcs,if (PCS_HAS_PINCONF && function) { res =pcs_parse_pinconf (pcs, np, function, map); -if (res) +if (res ==0 ) + *num_maps =2 ; +else if (res == -ENOTSUPP ) + *num_maps =1 ; +else goto free_pingroups; - *num_maps =2 ; }else { *num_maps =1 ; } --
Quoted 1 line
2.25.1