Skip to contentSkip to conversation

bitfield.h: add FIELD_MAX() and field_max()

PATCH Superseded Alex Elder 3 participants 5 messages Mar 6, 2020 – Mar 11, 2020 Bitmap API

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore

Alex Elder

author
#0 lore raw

PATCH v2 1/1 bitfield.h: add FIELD_MAX() and field_max()

Define FIELD_MAX(), which supplies the maximum value that can be represented by a bitfield. Define field_max() as well, to go along with the lower-case forms of the field mask functions.

---
v2: Added Acked-by from Jakub.

David, I added you to this because it's probably easiest to take this change in through your tree with the rest of the IPA code (which uses field_max(), defined here).

-Alex

 include/linux/bitfield.h | 14 ++++++++++++++
 1 file changed, 14 insertions(+)
include/linux/bitfield.h +14 -0
index 4bbb5f1c8b5b..48ea093ff04c 100644
@@ -55,6 +55,19 @@
 					      (1ULL << __bf_shf(_mask))); \
 	})
 
+/**
+ * FIELD_MAX() - produce the maximum value representable by a field
+ * @_mask: shifted mask defining the field's length and position
+ *
+ * FIELD_MAX() returns the maximum value that can be held in the field
+ * specified by @_mask.
+ */
+#define FIELD_MAX(_mask)						\
+	({								\
+		__BF_FIELD_CHECK(_mask, 0ULL, 0ULL, "FIELD_MAX: ");	\
+		(typeof(_mask))((_mask) >> __bf_shf(_mask));		\
+	})
+
 /**
  * FIELD_FIT() - check if value fits in the field
  * @_mask: shifted mask defining the field's length and position
@@ -110,6 +123,7 @@ static __always_inline u64 field_mask(u64 field)
 {
 	return field / field_multiplier(field);
 }
+#define field_max(field)	((typeof(field))field_mask(field))
 #define ____MAKE_OP(type,base,to,from)					\
 static __always_inline __##type type##_encode_bits(base v, base field)	\
 {									\
Signature · 1 line
2.20.1
#1 lore raw
Quoted 53 lines · Alex Elder

Define FIELD_MAX(), which supplies the maximum value that can be represented by a bitfield. Define field_max() as well, to go along with the lower-case forms of the field mask functions.

---
v2: Added Acked-by from Jakub.

David, I added you to this because it's probably easiest to take this change in through your tree with the rest of the IPA code (which uses field_max(), defined here).

-Alex

 include/linux/bitfield.h | 14 ++++++++++++++
 1 file changed, 14 insertions(+)
include/linux/bitfield.h +14 -0
index 4bbb5f1c8b5b..48ea093ff04c 100644
@@ -55,6 +55,19 @@
 					      (1ULL << __bf_shf(_mask))); \
 	})
 
+/**
+ * FIELD_MAX() - produce the maximum value representable by a field
+ * @_mask: shifted mask defining the field's length and position
+ *
+ * FIELD_MAX() returns the maximum value that can be held in the field
+ * specified by @_mask.
+ */
+#define FIELD_MAX(_mask)						\
+	({								\
+		__BF_FIELD_CHECK(_mask, 0ULL, 0ULL, "FIELD_MAX: ");	\
+		(typeof(_mask))((_mask) >> __bf_shf(_mask));		\
+	})
+
 /**
  * FIELD_FIT() - check if value fits in the field
  * @_mask: shifted mask defining the field's length and position
@@ -110,6 +123,7 @@ static __always_inline u64 field_mask(u64 field)
 {
 	return field / field_multiplier(field);
 }
+#define field_max(field)	((typeof(field))field_mask(field))
 #define ____MAKE_OP(type,base,to,from)					\
 static __always_inline __##type type##_encode_bits(base v, base field)	\
 {									\
Signature · 1 line
2.20.1

Without this patch, the IPA driver that was picked up a couple of days ago does not build...

$ make -j$(nproc) -s ARCH=arm64 CC=clang CROSS_COMPILE=aarch64-linux-gnu- O=out.arm32 distclean allyesconfig drivers/net/ipa/
../drivers/net/ipa/ipa_cmd.c:182:28: error: implicit declaration of function 'field_max' [-Werror,-Wimplicit-function-declaration]
        BUILD_BUG_ON(TABLE_SIZE > field_max(IP_FLTRT_FLAGS_HASH_SIZE_FMASK));
                                  ^
../drivers/net/ipa/ipa_cmd.c:182:28: note: did you mean 'field_mask'?
../include/linux/bitfield.h:109:28: note: 'field_mask' declared here
static __always_inline u64 field_mask(u64 field)
                           ^
../drivers/net/ipa/ipa_cmd.c:183:28: error: implicit declaration of function 'field_max' [-Werror,-Wimplicit-function-declaration]
        BUILD_BUG_ON(TABLE_SIZE > field_max(IP_FLTRT_FLAGS_NHASH_SIZE_FMASK));
                                  ^
../drivers/net/ipa/gsi.c:220:39: error: implicit declaration of function 'field_max' [-Werror,-Wimplicit-function-declaration]
        BUILD_BUG_ON(GSI_RING_ELEMENT_SIZE > field_max(ELEMENT_SIZE_FMASK));
                                             ^
2 errors generated.
../drivers/net/ipa/gsi.c:220:39: note: did you mean 'field_mask'?
../include/linux/bitfield.h:109:28: note: 'field_mask' declared here
static __always_inline u64 field_mask(u64 field)
                           ^
../drivers/net/ipa/gsi.c:223:39: error: implicit declaration of function 'field_max' [-Werror,-Wimplicit-function-declaration]
        BUILD_BUG_ON(GSI_RING_ELEMENT_SIZE > field_max(EV_ELEMENT_SIZE_FMASK));
                                             ^
make[5]: *** [../scripts/Makefile.build:267: drivers/net/ipa/ipa_cmd.o] Error 1
make[5]: *** Waiting for unfinished jobs....
../drivers/net/ipa/gsi.c:710:16: error: implicit declaration of function 'field_max' [-Werror,-Wimplicit-function-declaration]
                wrr_weight = field_max(WRR_WEIGHT_FMASK);
                             ^
3 errors generated.
make[5]: *** [../scripts/Makefile.build:267: drivers/net/ipa/gsi.o] Error 1
../drivers/net/ipa/ipa_endpoint.c:584:14: error: implicit declaration of function 'field_max' [-Werror,-Wimplicit-function-declaration]
        if (scale > field_max(SCALE_FMASK))
                    ^
../drivers/net/ipa/ipa_endpoint.c:584:14: note: did you mean 'field_mask'?
../include/linux/bitfield.h:109:28: note: 'field_mask' declared here
static __always_inline u64 field_mask(u64 field)
                           ^
../drivers/net/ipa/ipa_endpoint.c:965:16: error: implicit declaration of function 'field_max' [-Werror,-Wimplicit-function-declaration]
        return val == field_max(IPA_STATUS_FLAGS1_RT_RULE_ID_FMASK);
                      ^
2 errors generated.

This probably should go through the net-next tree to avoid that build breakage.

Cheers,
Nathan

#2 lore raw
Quoted 2 lines · On Tue, 10 Mar 2020 14:29:38 -0700 Nathan Chancellor wrote:

Without this patch, the IPA driver that was picked up a couple of days ago does not build...

😳

Yes please, Alex could you repost ASAP with [PATCH net-next] subject and CC netdev to get it into the netdev patchwork?

Please also make IPA:

	depends on (ARCH_QCOM || COMPILE_TEST) && 64BIT && NET

Otherwise it's really hard to make sure the code builds.

author
#3 lore raw
Quoted 6 lines · Jakub Kicinski
Quoted 2 lines · On Tue, 10 Mar 2020 14:29:38 -0700 Nathan Chancellor wrote:

Without this patch, the IPA driver that was picked up a couple of days ago does not build...

😳

Quoted 8 lines

Yes please, Alex could you repost ASAP with [PATCH net-next] subject and CC netdev to get it into the netdev patchwork?

Please also make IPA:

	depends on (ARCH_QCOM || COMPILE_TEST) && 64BIT && NET

Otherwise it's really hard to make sure the code builds.

Sorry all. I have been on vacation the last few days and only now saw this.

I will put this together shortly.

-Alex

author
#4 lore raw
Quoted 12 lines · Alex Elder
Quoted 11 lines · Jakub Kicinski
Quoted 2 lines · On Tue, 10 Mar 2020 14:29:38 -0700 Nathan Chancellor wrote:

Without this patch, the IPA driver that was picked up a couple of days ago does not build...

😳

Yes please, Alex could you repost ASAP with [PATCH net-next] subject and CC netdev to get it into the netdev patchwork?

Please also make IPA:

Quoted 8 lines
Quoted 3 lines
	depends on (ARCH_QCOM || COMPILE_TEST) && 64BIT && NET

Otherwise it's really hard to make sure the code builds.

Sorry all. I have been on vacation the last few days and only now saw this.

I will put this together shortly.

I just sent an updated version of the field_max() patch. I will do the COMPILE_TEST change separately (and not today).

-Alex

Quoted 2 lines

-Alex