Skip to contentSkip to conversation

Introduce the for_each_set_clump macro

PATCH Superseded Syed Nayyar Waris 4 participants 12 messages Oct 6, 2020 – Oct 23, 2020

Subsystems (8)

v11 of 12 Superseded Latest is v12 →

lore
Share & tools
lore
#0 lore raw

v11 cover Introduce the for_each_set_clump macro

Hello Linus,

Since this patchset primarily affects GPIO drivers, would you like to pick it up through your GPIO tree?

This patchset introduces a new generic version of for_each_set_clump. The previous version of for_each_set_clump8 used a fixed size 8-bit clump, but the new generic version can work with clump of any size but less than or equal to BITS_PER_LONG. The patchset utilizes the new macro in several GPIO drivers.

The earlier 8-bit for_each_set_clump8 facilitated a
for-loop syntax that iterates over a memory region entire groups of set bits at a time.

For example, suppose you would like to iterate over a 32-bit integer 8 bits at a time, skipping over 8-bit groups with no set bit, where XXXXXXXX represents the current 8-bit group:

    Example:        10111110 00000000 11111111 00110011
    First loop:     10111110 00000000 11111111 XXXXXXXX
    Second loop:    10111110 00000000 XXXXXXXX 00110011
    Third loop:     XXXXXXXX 00000000 11111111 00110011

Each iteration of the loop returns the next 8-bit group that has at least one set bit.

But with the new for_each_set_clump the clump size can be different from 8 bits. Moreover, the clump can be split at word boundary in situations where word size is not multiple of clump size. Following are examples showing the working of new macro for clump sizes of 24 bits and 6 bits.

Example 1:
clump size: 24 bits, Number of clumps (or ports): 10
bitmap stores the bit information from where successive clumps are retrieved.

     /* bitmap memory region */
        0x00aa0000ff000000;  /* Most significant bits */
        0xaaaaaa0000ff0000;
        0x000000aa000000aa;
        0xbbbbabcdeffedcba;  /* Least significant bits */

Different iterations of for_each_set_clump:-
'offset' is the bit position and 'clump' is the 24 bit clump from the above bitmap.

Iteration first:        offset: 0 clump: 0xfedcba
Iteration second:       offset: 24 clump: 0xabcdef
Iteration third:        offset: 48 clump: 0xaabbbb
Iteration fourth:       offset: 96 clump: 0xaa
Iteration fifth:        offset: 144 clump: 0xff
Iteration sixth:        offset: 168 clump: 0xaaaaaa
Iteration seventh:      offset: 216 clump: 0xff

Loop breaks because in the end the remaining bits (0x00aa) size was less than clump size of 24 bits.

In above example it can be seen that in iteration third, the 24 bit clump that was retrieved was split between bitmap[0] and bitmap[1]. This example also shows that 24 bit zeroes if present in between, were skipped (preserving the previous for_each_set_macro8 behaviour).

Example 2:
clump size = 6 bits, Number of clumps (or ports) = 3.

     /* bitmap memory region */
        0x00aa0000ff000000;  /* Most significant bits */
        0xaaaaaa0000ff0000;
        0x0f00000000000000;
        0x0000000000000ac0;  /* Least significant bits */

Different iterations of for_each_set_clump:
'offset' is the bit position and 'clump' is the 6 bit clump from the above bitmap.
Iteration first: offset: 6 clump: 0x2b
Loop breaks because 6 * 3 = 18 bits traversed in bitmap. Here 6 * 3 is clump size * no. of clumps.

Changes in v11:
- [Patch 1/4]: Document range of values 'nbits' can take.
- [Patch 4/4]: Change variable name 'flag' to 'flags'.

Changes in v10:
- Patchset based on v5.9-rc1.

Changes in v9:
- [Patch 4/4]: Remove looping of 'for_each_set_clump' and instead process two halves of a 64-bit bitmap separately or individually. Use normal spin_lock call for second inner lock. And take the spin_lock_init call outside the 'if' condition in the probe function of driver.

Changes in v8:
- [Patch 2/4]: Minor change: Use '__initdata' for correct section mismatch in 'clump_test_data' array.

Changes in v7:
- [Patch 2/4]: Minor changes: Use macro 'DECLARE_BITMAP()' and split 'struct' definition and test data.

Changes in v6:
- [Patch 2/4]: Make 'for loop' inside test_for_each_set_clump more succinct.

Changes in v5:
- [Patch 4/4]: Minor change: Hardcode value for better code readability.

Changes in v4:
- [Patch 2/4]: Use 'for' loop in test function of for_each_set_clump.
- [Patch 3/4]: Minor change: Inline value for better code readability.
- [Patch 4/4]: Minor change: Inline value for better code readability.

Changes in v3:
- [Patch 3/4]: Change datatype of some variables from u64 to unsigned long in function thunderx_gpio_set_multiple.

CHanges in v2:
- [Patch 2/4]: Unify different tests for 'for_each_set_clump'. Pass test data as function parameters.
- [Patch 2/4]: Remove unnecessary bitmap_zero calls.

Syed Nayyar Waris (4):
bitops: Introduce the for_each_set_clump macro lib/test_bitmap.c: Add for_each_set_clump test cases
gpio: thunderx: Utilize for_each_set_clump macro
gpio: xilinx: Utilize generic bitmap_get_value and _set_value

 drivers/gpio/gpio-thunderx.c      |  11 ++-
 drivers/gpio/gpio-xilinx.c        |  64 ++++++-------
 include/asm-generic/bitops/find.h |  19 ++++
 include/linux/bitmap.h            |  63 +++++++++++++
 include/linux/bitops.h            |  13 +++
 lib/find_bit.c                    |  14 +++
 lib/test_bitmap.c                 | 144 ++++++++++++++++++++++++++++++
 7 files changed, 292 insertions(+), 36 deletions(-)

base-commit: 9123e3a74ec7b934a4a099e98af6a61c2f80bbf5

Signature · 1 line
2.26.2
#1 lore raw

PATCH v11 1/4 bitops: Introduce the for_each_set_clump macro

This macro iterates for each group of bits (clump) with set bits, within a bitmap memory region. For each iteration, "start" is set to the bit offset of the found clump, while the respective clump value is stored to the location pointed by "clump". Additionally, the bitmap_get_value() and bitmap_set_value() functions are introduced to respectively get and set a value of n-bits in a bitmap memory region. The n-bits can have any size less than or equal to BITS_PER_LONG. Moreover, during setting value of n-bit in bitmap, if a situation arise that the width of next n-bit is exceeding the word boundary, then it will divide itself such that some portion of it is stored in that word, while the remaining portion is stored in the next higher word. Similar situation occurs while retrieving the value from bitmap.

---
Changes in v11:
- Document valid range of values that 'nbits' can take.

Changes in v10:
- No change.

Changes in v9:
- No change.

Changes in v8:
- No change.

Changes in v7:
- No change.

Changes in v6:
- No change.

Changes in v5:
- No change.

Changes in v4:
- No change.

Changes in v3:
- No change.

Changes in v2:
- No change.

 include/asm-generic/bitops/find.h | 19 ++++++++++
 include/linux/bitmap.h            | 63 +++++++++++++++++++++++++++++++
 include/linux/bitops.h            | 13 +++++++
 lib/find_bit.c                    | 14 +++++++
 4 files changed, 109 insertions(+)
include/asm-generic/bitops/find.h +19 -0
index 9fdf21302fdf..4e6600759455 100644
@@ -97,4 +97,23 @@ extern unsigned long find_next_clump8(unsigned long *clump,
 #define find_first_clump8(clump, bits, size) \
 	find_next_clump8((clump), (bits), (size), 0)
 
+/**
+ * find_next_clump - find next clump with set bits in a memory region
+ * @clump: location to store copy of found clump
+ * @addr: address to base the search on
+ * @size: bitmap size in number of bits
+ * @offset: bit offset at which to start searching
+ * @clump_size: clump size in bits
+ *
+ * Returns the bit offset for the next set clump; the found clump value is
+ * copied to the location pointed by @clump. If no bits are set, returns @size.
+ */
+extern unsigned long find_next_clump(unsigned long *clump,
+				      const unsigned long *addr,
+				      unsigned long size, unsigned long offset,
+				      unsigned long clump_size);
+
+#define find_first_clump(clump, bits, size, clump_size) \
+	find_next_clump((clump), (bits), (size), 0, (clump_size))
+
 #endif /*_ASM_GENERIC_BITOPS_FIND_H_ */
include/linux/bitmap.h +63 -0

82 lines not on this page · raw message

include/linux/bitops.h +13 -0
index 99f2ac30b1d9..36a445e4a7cc 100644
@@ -62,6 +62,19 @@ extern unsigned long __sw_hweight64(__u64 w);
 	     (start) < (size); \
 	     (start) = find_next_clump8(&(clump), (bits), (size), (start) + 8))
 
+/**
+ * for_each_set_clump - iterate over bitmap for each clump with set bits
+ * @start: bit offset to start search and to store the current iteration offset
+ * @clump: location to store copy of current 8-bit clump
+ * @bits: bitmap address to base the search on
+ * @size: bitmap size in number of bits
+ * @clump_size: clump size in bits
+ */
+#define for_each_set_clump(start, clump, bits, size, clump_size) \
+	for ((start) = find_first_clump(&(clump), (bits), (size), (clump_size)); \
+	     (start) < (size); \
+	     (start) = find_next_clump(&(clump), (bits), (size), (start) + (clump_size), (clump_size)))
+
 static inline int get_bitmask_order(unsigned int count)
 {
 	int order;
lib/find_bit.c +14 -0
index 49f875f1baf7..1341bd39b32a 100644
@@ -190,3 +190,17 @@ unsigned long find_next_clump8(unsigned long *clump, const unsigned long *addr,
 	return offset;
 }
 EXPORT_SYMBOL(find_next_clump8);
+
+unsigned long find_next_clump(unsigned long *clump, const unsigned long *addr,
+			       unsigned long size, unsigned long offset,
+			       unsigned long clump_size)
+{
+	offset = find_next_bit(addr, size, offset);
+	if (offset == size)
+		return size;
+
+	offset = rounddown(offset, clump_size);
+	*clump = bitmap_get_value(addr, offset, clump_size);
+	return offset;
+}
+EXPORT_SYMBOL(find_next_clump);
Signature · 1 line
2.26.2
#2 lore raw

Re: [PATCH v11 1/4] bitops: Introduce the for_each_set_clump macro

Quoted 12 lines · Syed Nayyar Waris

This macro iterates for each group of bits (clump) with set bits, within a bitmap memory region. For each iteration, "start" is set to the bit offset of the found clump, while the respective clump value is stored to the location pointed by "clump". Additionally, the bitmap_get_value() and bitmap_set_value() functions are introduced to respectively get and set a value of n-bits in a bitmap memory region. The n-bits can have any size less than or equal to BITS_PER_LONG. Moreover, during setting value of n-bit in bitmap, if a situation arise that the width of next n-bit is exceeding the word boundary, then it will divide itself such that some portion of it is stored in that word, while the remaining portion is stored in the next higher word. Similar situation occurs while retrieving the value from bitmap.

...

include/linux/bitmap.h line 82

Quoted 9 lines
include/linux/bitmap.h +4 -0
@@ -75,7 +75,11 @@
  *  bitmap_from_arr32(dst, buf, nbits)          Copy nbits from u32[] buf to dst
  *  bitmap_to_arr32(buf, src, nbits)            Copy nbits from buf to u32[] dst
  *  bitmap_get_value8(map, start)               Get 8bit value from map at start
+ *  bitmap_get_value(map, start, nbits)		Get bit value of size
+ *						'nbits' from map at start
  *  bitmap_set_value8(map, value, start)        Set 8bit value to map at start
+ *  bitmap_set_value(map, value, start, nbits)	Set bit value of size 'nbits'
+ *						of map at start

Formatting here is done with solely spaces, no TABs.

...

include/linux/bitmap.h static inline unsigned long bitmap_get_value8(const unsigned long *map, · line 575

Quoted 8 lines
include/linux/bitmap.h +6 -0
+/**
+ * bitmap_get_value - get a value of n-bits from the memory region
+ * @map: address to the bitmap memory region
+ * @start: bit offset of the n-bit value
+ * @nbits: size of value in bits (must be between 1 and BITS_PER_LONG inclusive).
 
 
+ *	nbits less than 1 or more than BITS_PER_LONG causes undefined behaviour.

Please, detach this from field description and move to a main description.

Quoted 4 lines

+ *
+ * Returns value of nbits located at the @start bit offset within the @map
+ * memory region.
+ */

...

include/linux/bitmap.h static inline unsigned long bitmap_get_value8(const unsigned long *map, · line 591

Quoted 1 line
include/linux/bitmap.h +1 -0
+		return (map[index] >> offset) & GENMASK(nbits - 1, 0);

Have you considered to use rather BIT{_ULL}(nbits) - 1? It maybe better for code generation.

...

Quoted 8 lines

+/**
+ * bitmap_set_value - set n-bit value within a memory region
+ * @map: address to the bitmap memory region
+ * @value: value of nbits
+ * @start: bit offset of the n-bit value
+ * @nbits: size of value in bits (must be between 1 and BITS_PER_LONG inclusive).

+ * nbits less than 1 or more than BITS_PER_LONG causes undefined behaviour.

Please, detach this from field description and move to a main description.

Quoted 1 line

+ */

...

include/linux/bitmap.h static inline void bitmap_set_value8(unsigned long *map, unsigned long value, · line 632

Quoted 1 line
include/linux/bitmap.h +1 -0
+	value &= GENMASK(nbits - 1, 0);

Ditto.

include/linux/bitmap.h static inline void bitmap_set_value8(unsigned long *map, unsigned long value, · line 635

Quoted 1 line
include/linux/bitmap.h +1 -0
+		map[index] &= ~(GENMASK(nbits + offset - 1, offset));

Last time I checked such GENMASK) use, it gave a lot of code when GENMASK(nbits - 1, 0) << offset works much better, but see also above.

Signature · 2 lines
With Best Regards,
Andy Shevchenko
#3 lore raw

Re: [PATCH v11 1/4] bitops: Introduce the for_each_set_clump macro

Quoted 17 lines · Andy Shevchenko
Quoted 12 lines · Syed Nayyar Waris

This macro iterates for each group of bits (clump) with set bits, within a bitmap memory region. For each iteration, "start" is set to the bit offset of the found clump, while the respective clump value is stored to the location pointed by "clump". Additionally, the bitmap_get_value() and bitmap_set_value() functions are introduced to respectively get and set a value of n-bits in a bitmap memory region. The n-bits can have any size less than or equal to BITS_PER_LONG. Moreover, during setting value of n-bit in bitmap, if a situation arise that the width of next n-bit is exceeding the word boundary, then it will divide itself such that some portion of it is stored in that word, while the remaining portion is stored in the next higher word. Similar situation occurs while retrieving the value from bitmap.

...

Quoted 11 lines

include/linux/bitmap.h line 82

Quoted 9 lines
include/linux/bitmap.h +4 -0
@@ -75,7 +75,11 @@
  *  bitmap_from_arr32(dst, buf, nbits)          Copy nbits from u32[] buf to dst
  *  bitmap_to_arr32(buf, src, nbits)            Copy nbits from buf to u32[] dst
  *  bitmap_get_value8(map, start)               Get 8bit value from map at start
+ *  bitmap_get_value(map, start, nbits)              Get bit value of size
+ *                                           'nbits' from map at start
  *  bitmap_set_value8(map, value, start)        Set 8bit value to map at start
+ *  bitmap_set_value(map, value, start, nbits)       Set bit value of size 'nbits'
+ *                                           of map at start

Formatting here is done with solely spaces, no TABs.

Okay. Done

Quoted 10 lines

...

Quoted 5 lines

+/**
+ * bitmap_get_value - get a value of n-bits from the memory region
+ * @map: address to the bitmap memory region
+ * @start: bit offset of the n-bit value
+ * @nbits: size of value in bits (must be between 1 and BITS_PER_LONG inclusive).

Quoted 3 lines
Quoted 1 line

+ * nbits less than 1 or more than BITS_PER_LONG causes undefined behaviour.

Please, detach this from field description and move to a main description.

Okay. Done.

Quoted 6 lines
Quoted 5 lines

+ *
+ * Returns value of nbits located at the @start bit offset within the @map
+ * memory region.
+ */

Quoted 6 lines

...

Quoted 1 line

+ return (map[index] >> offset) & GENMASK(nbits - 1, 0);

Have you considered to use rather BIT{_ULL}(nbits) - 1? It maybe better for code generation.

Yes I have considered using BIT{_ULL} in earlier versions of patchset. It has a problem:

This macro when used in both bitmap_get_value and
bitmap_set_value functions, it will give unexpected results when nbits or clump size is BITS_PER_LONG (32 or 64 depending on arch).

Actually when nbits (clump size) is 64 (BITS_PER_LONG is 64, for example), (BIT(nbits) - 1)
gives a value of zero and when this zero is ANDed with any value, it makes it full zero. This is unexpected, and incorrect calculation occurs.

What actually happens is in the macro expansion of BIT(64), that is 1 << 64, the '1' overflows from leftmost bit position (most significant bit) and re-enters at the rightmost bit position (least significant bit), therefore 1 << 64 becomes '0x1', and when another '1' is subtracted from this, the final result becomes 0.

This is undefined behavior in the C standard (section 6.5.7 in the N1124)

Quoted 10 lines

...

Quoted 6 lines

+/**
+ * bitmap_set_value - set n-bit value within a memory region
+ * @map: address to the bitmap memory region
+ * @value: value of nbits
+ * @start: bit offset of the n-bit value
+ * @nbits: size of value in bits (must be between 1 and BITS_PER_LONG inclusive).

Quoted 3 lines
Quoted 1 line

+ * nbits less than 1 or more than BITS_PER_LONG causes undefined behaviour.

Please, detach this from field description and move to a main description.

Okay. Done

Quoted 5 lines
Quoted 2 lines

+ */

...

Quoted 8 lines
Quoted 1 line

+ value &= GENMASK(nbits - 1, 0);

Ditto.

Quoted 1 line

+ map[index] &= ~(GENMASK(nbits + offset - 1, offset));

Last time I checked such GENMASK) use, it gave a lot of code when GENMASK(nbits - 1, 0) << offset works much better, but see also above.

Yes I have incorporated your suggestion to use the '<<' operator. Thank You.

Quoted 6 lines

--
With Best Regards,
Andy Shevchenko

#4 lore raw

Re: [PATCH v11 1/4] bitops: Introduce the for_each_set_clump macro

Quoted 3 lines · Syed Nayyar Waris
Quoted 1 line · Andy Shevchenko

On Tue, Oct 06, 2020 at 02:52:16PM +0530, Syed Nayyar Waris wrote:

...

Quoted 17 lines
Quoted 4 lines
Quoted 1 line

+ return (map[index] >> offset) & GENMASK(nbits - 1, 0);

Have you considered to use rather BIT{_ULL}(nbits) - 1? It maybe better for code generation.

Yes I have considered using BIT{_ULL} in earlier versions of patchset. It has a problem:

This macro when used in both bitmap_get_value and
bitmap_set_value functions, it will give unexpected results when nbits or clump size is BITS_PER_LONG (32 or 64 depending on arch).

Actually when nbits (clump size) is 64 (BITS_PER_LONG is 64, for example), (BIT(nbits) - 1)
gives a value of zero and when this zero is ANDed with any value, it makes it full zero. This is unexpected, and incorrect calculation occurs.

Quoted 7 lines

What actually happens is in the macro expansion of BIT(64), that is 1 << 64, the '1' overflows from leftmost bit position (most significant bit) and re-enters at the rightmost bit position (least significant bit), therefore 1 << 64 becomes '0x1', and when another '1' is subtracted from this, the final result becomes 0.

This is undefined behavior in the C standard (section 6.5.7 in the N1124)

I see, indeed, for 64/32 it is like this.

...

Quoted 1 line

Yes I have incorporated your suggestion to use the '<<' operator. Thank You.

One side note, consider the use round_up() vs. roundup(). I don't remember which one is optimized to divisor being power of 2.

Signature · 2 lines
With Best Regards,
Andy Shevchenko
#5 lore raw

Re: [PATCH v11 1/4] bitops: Introduce the for_each_set_clump macro

Quoted 33 lines · Andy Shevchenko
Quoted 3 lines · Syed Nayyar Waris
Quoted 1 line · Andy Shevchenko

On Tue, Oct 06, 2020 at 02:52:16PM +0530, Syed Nayyar Waris wrote:

...

Quoted 17 lines
Quoted 4 lines
Quoted 1 line

+ return (map[index] >> offset) & GENMASK(nbits - 1, 0);

Have you considered to use rather BIT{_ULL}(nbits) - 1? It maybe better for code generation.

Yes I have considered using BIT{_ULL} in earlier versions of patchset. It has a problem:

This macro when used in both bitmap_get_value and
bitmap_set_value functions, it will give unexpected results when nbits or clump size is BITS_PER_LONG (32 or 64 depending on arch).

Actually when nbits (clump size) is 64 (BITS_PER_LONG is 64, for example), (BIT(nbits) - 1)
gives a value of zero and when this zero is ANDed with any value, it makes it full zero. This is unexpected, and incorrect calculation occurs.

Quoted 7 lines

What actually happens is in the macro expansion of BIT(64), that is 1 << 64, the '1' overflows from leftmost bit position (most significant bit) and re-enters at the rightmost bit position (least significant bit), therefore 1 << 64 becomes '0x1', and when another '1' is subtracted from this, the final result becomes 0.

This is undefined behavior in the C standard (section 6.5.7 in the N1124)

Quoted 8 lines

I see, indeed, for 64/32 it is like this.

...

Quoted 1 line

Yes I have incorporated your suggestion to use the '<<' operator. Thank You.

One side note, consider the use round_up() vs. roundup(). I don't remember which one is optimized to divisor being power of 2.

Yes. changed 'roundup' to 'round_up'. 'round_up' is optimized for power-of-2. Thank you.

Syed Nayyar Waris

#6 lore raw

PATCH v11 2/4 lib/test_bitmap.c: Add for_each_set_clump test cases

In reply to Syed Nayyar Waris

The introduction of the generic for_each_set_clump macro need test cases to verify the implementation. This patch adds test cases for scenarios in which clump sizes are 8 bits, 24 bits, 30 bits and 6 bits. The cases contain situations where clump is getting split at the word boundary and also when zeroes are present in the start and middle of bitmap.

---
Changes in v11:
- No change.

Changes in v10:
- No change.

Changes in v9:
- No change.

Changes in v8:
- [Patch 2/4]: Minor change: Use '__initdata' for correct section mismatch in 'clump_test_data' array.

Changes in v7:
- Minor changes: Use macro 'DECLARE_BITMAP()' and split 'struct' definition and test data.

Changes in v6:
- Make 'for loop' inside 'test_for_each_set_clump' more succinct.

Changes in v5:
- No change.

Changes in v4:
- Use 'for' loop in test function of 'for_each_set_clump'.

Changes in v3:
- No Change.

Changes in v2:
- Unify different tests for 'for_each_set_clump'. Pass test data as function parameters.
- Remove unnecessary bitmap_zero calls.

 lib/test_bitmap.c | 144 ++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 144 insertions(+)
lib/test_bitmap.c +144 -0
index df903c53952b..cb2cf3858f93 100644
@@ -155,6 +155,37 @@ static bool __init __check_eq_clump8(const char *srcfile, unsigned int line,
 	return true;
 }
 
+static bool __init __check_eq_clump(const char *srcfile, unsigned int line,
+				    const unsigned int offset,
+				    const unsigned int size,
+				    const unsigned long *const clump_exp,
+				    const unsigned long *const clump,
+				    const unsigned long clump_size)
+{
+	unsigned long exp;
+
+	if (offset >= size) {
+		pr_warn("[%s:%u] bit offset for clump out-of-bounds: expected less than %u, got %u\n",
+			srcfile, line, size, offset);
+		return false;
+	}
+
+	exp = clump_exp[offset / clump_size];
+	if (!exp) {
+		pr_warn("[%s:%u] bit offset for zero clump: expected nonzero clump, got bit offset %u with clump value 0",
+			srcfile, line, offset);
+		return false;
+	}
+
+	if (*clump != exp) {
+		pr_warn("[%s:%u] expected clump value of 0x%lX, got clump value of 0x%lX",
+			srcfile, line, exp, *clump);
+		return false;
+	}
+
+	return true;
+}
+
 #define __expect_eq(suffix, ...)					\
 	({								\
 		int result = 0;						\
@@ -172,6 +203,7 @@ static bool __init __check_eq_clump8(const char *srcfile, unsigned int line,
 #define expect_eq_pbl(...)		__expect_eq(pbl, ##__VA_ARGS__)
 #define expect_eq_u32_array(...)	__expect_eq(u32_array, ##__VA_ARGS__)
 #define expect_eq_clump8(...)		__expect_eq(clump8, ##__VA_ARGS__)
+#define expect_eq_clump(...)		__expect_eq(clump, ##__VA_ARGS__)
 
 static void __init test_zero_clear(void)
 {
@@ -577,6 +609,28 @@ static void noinline __init test_mem_optimisations(void)
 	}
 }
 
+static const unsigned long clump_bitmap_data[] __initconst = {
+	0x38000201,
+	0x05ff0f38,
+	0xeffedcba,
+	0xbbbbabcd,
+	0x000000aa,
+	0x000000aa,
+	0x00ff0000,
+	0xaaaaaa00,
+	0xff000000,
+	0x00aa0000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x0f000000,
+	0x00ff0000,
+	0xaaaaaa00,
+	0xff000000,
+	0x00aa0000,
+	0x00000ac0,
+};
+
 static const unsigned char clump_exp[] __initconst = {
 	0x01,	/* 1 bit set */
 	0x02,	/* non-edge 1 bit set */
@@ -588,6 +642,95 @@ static const unsigned char clump_exp[] __initconst = {
 	0x05,	/* non-adjacent 2 bits set */
 };
 
+static const unsigned long clump_exp1[] __initconst = {
+	0x01,	/* 1 bit set */
+	0x02,	/* non-edge 1 bit set */
+	0x00,	/* zero bits set */
+	0x38,	/* 3 bits set across 4-bit boundary */
+	0x38,	/* Repeated clump */
+	0x0F,	/* 4 bits set */
+	0xFF,	/* all bits set */
+	0x05,	/* non-adjacent 2 bits set */
+};
+
+static const unsigned long clump_exp2[] __initconst = {
+	0xfedcba,	/* 24 bits */
+	0xabcdef,
+	0xaabbbb,	/* Clump split between 2 words */
+	0x000000,	/* zeroes in between */
+	0x0000aa,
+	0x000000,
+	0x0000ff,
+	0xaaaaaa,
+	0x000000,
+	0x0000ff,
+};
+
+static const unsigned long clump_exp3[] __initconst = {
+	0x00000000,	/* starting with 0s*/
+	0x00000000,	/* All 0s */
+	0x00000000,
+	0x00000000,
+	0x3f00000f,     /* Non zero set */
+	0x2aa80003,
+	0x00000aaa,
+	0x00003fc0,
+};
+
+static const unsigned long clump_exp4[] __initconst = {
+	0x00,
+	0x2b,
+};
+
+struct clump_test_data_params {
+	DECLARE_BITMAP(data, 256);
+	unsigned long count;
+	unsigned long offset;
+	unsigned long limit;
+	unsigned long clump_size;
+	unsigned long const *exp;
+};
+
+static struct clump_test_data_params clump_test_data[] __initdata =
+				      { {{0}, 2, 0, 64, 8, clump_exp1},
+					{{0}, 8, 2, 240, 24, clump_exp2},
+					{{0}, 8, 10, 240, 30, clump_exp3},
+					{{0}, 1, 18, 18, 6, clump_exp4} };
+
+static void __init prepare_test_data(unsigned int index)
+{
+	int i;
+	unsigned long width = 0;
+
+	for(i = 0; i < clump_test_data[index].count; i++)
+	{
+		bitmap_set_value(clump_test_data[index].data,
+			clump_bitmap_data[(clump_test_data[index].offset)++], width, 32);
+		width += 32;
+	}
+}
+
+static void __init execute_for_each_set_clump_test(unsigned int index)
+{
+	unsigned long start, clump;
+
+	for_each_set_clump(start, clump, clump_test_data[index].data,
+						clump_test_data[index].limit,
+						clump_test_data[index].clump_size)
+	expect_eq_clump(start, clump_test_data[index].limit, clump_test_data[index].exp,
+						&clump, clump_test_data[index].clump_size);
+}
+
+static void __init test_for_each_set_clump(void)
+{
+	unsigned int i;
+
+	for (i = 0; i < ARRAY_SIZE(clump_test_data); i++) {
+		prepare_test_data(i);
+		execute_for_each_set_clump_test(i);
+	}
+}
+
 static void __init test_for_each_set_clump8(void)
 {
 #define CLUMP_EXP_NUMBITS 64
@@ -680,6 +823,7 @@ static void __init selftest(void)
 	test_bitmap_parselist_user();
 	test_mem_optimisations();
 	test_for_each_set_clump8();
+	test_for_each_set_clump();
 	test_bitmap_cut();
 }
 
Signature · 1 line
2.26.2
#7 lore raw

Re: [PATCH v11 2/4] lib/test_bitmap.c: Add for_each_set_clump test cases

I love your patch! Perhaps something to improve:

Hi Syed,

I love your patch! Perhaps something to improve:

[auto build test WARNING on 9123e3a74ec7b934a4a099e98af6a61c2f80bbf5]

url:    https://github.com/0day-ci/linux/commits/Syed-Nayyar-Waris/Introduce-the-for_each_set_clump-macro/20201006-172813
base:    9123e3a74ec7b934a4a099e98af6a61c2f80bbf5
config: xtensa-randconfig-r026-20201005 (attached as .config)
compiler: xtensa-linux-gcc (GCC) 9.3.0

reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# https://github.com/0day-ci/linux/commit/bdf8587767d3df86c8c6fbc039089bf66bde631f
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Syed-Nayyar-Waris/Introduce-the-for_each_set_clump-macro/20201006-172813 git checkout bdf8587767d3df86c8c6fbc039089bf66bde631f
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=xtensa

If you fix the issue, kindly add following tag as appropriate

All warnings (new ones prefixed by >>, old ones prefixed by <<):

<< WARNING: modpost: lib/test_bitmap.o(.text.unlikely+0x1c): Section mismatch in reference from the function bitmap_equal() to the variable .init.rodata:clump_exp

>> WARNING: modpost: lib/test_bitmap.o(.text.unlikely+0x1c): Section mismatch in reference from the function bitmap_equal() to the variable .init.rodata:clump_exp4

The function bitmap_equal() references
the variable __initconst clump_exp4.
This is often because bitmap_equal lacks a __initconst annotation or the annotation of clump_exp4 is wrong.

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org

#8 lore raw

PATCH v11 3/4 gpio: thunderx: Utilize for_each_set_clump macro

In reply to Syed Nayyar Waris

This patch reimplements the thunderx_gpio_set_multiple function in drivers/gpio/gpio-thunderx.c to use the new for_each_set_clump macro. Instead of looping for each bank in thunderx_gpio_set_multiple function, now we can skip bank which is not set and save cycles.

---
Changes in v11:
- No change.

Changes in v10:
- No change.

Changes in v9:
- No change.

Changes in v8:
- No change.

Changes in v7:
- No change.

Changes in v6:
- No change.

Changes in v5:
- No change.

Changes in v4:
- Minor change: Inline value '64' in code for better code readability.

Changes in v3:
- Change datatype of some variables from u64 to unsigned long in function thunderx_gpio_set_multiple.

Changes in v2:
- No change.

 drivers/gpio/gpio-thunderx.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)
drivers/gpio/gpio-thunderx.c +7 -4
index 9f66deab46ea..58c9bb25a377 100644
@@ -275,12 +275,15 @@ static void thunderx_gpio_set_multiple(struct gpio_chip *chip,
 				       unsigned long *bits)
 {
 	int bank;
-	u64 set_bits, clear_bits;
+	unsigned long set_bits, clear_bits, gpio_mask;
+	unsigned long offset;
+
 	struct thunderx_gpio *txgpio = gpiochip_get_data(chip);
 
-	for (bank = 0; bank <= chip->ngpio / 64; bank++) {
-		set_bits = bits[bank] & mask[bank];
-		clear_bits = ~bits[bank] & mask[bank];
+	for_each_set_clump(offset, gpio_mask, mask, chip->ngpio, 64) {
+		bank = offset / 64;
+		set_bits = bits[bank] & gpio_mask;
+		clear_bits = ~bits[bank] & gpio_mask;
 		writeq(set_bits, txgpio->register_base + (bank * GPIO_2ND_BANK) + GPIO_TX_SET);
 		writeq(clear_bits, txgpio->register_base + (bank * GPIO_2ND_BANK) + GPIO_TX_CLR);
 	}
Signature · 1 line
2.26.2
#9 lore raw

PATCH v11 4/4 gpio: xilinx: Utilize generic bitmap_get_value and _set_value

In reply to Syed Nayyar Waris

This patch reimplements the xgpio_set_multiple function in drivers/gpio/gpio-xilinx.c to use the new generic functions:
bitmap_get_value and bitmap_set_value. The code is now simpler to read and understand. Moreover, instead of looping for each bit in xgpio_set_multiple function, now we can check each channel at a time and save cycles.

---
Changes in v11:
- Change variable name 'flag' to 'flags'.

Changes in v10:
- No change.

Changes in v9:
- Remove looping of 'for_each_set_clump' and instead process two halves of a 64-bit bitmap separately or individually. Use normal spin_lock call for second inner lock. And take the spin_lock_init call outside the 'if' condition in the 'probe' function of driver.

Changes in v8:
- No change.

Changes in v7:
- No change.

Changes in v6:
- No change.

Changes in v5:
- Minor change: Inline values '32' and '64' in code for better code readability.

Changes in v4:
- Minor change: Inline values '32' and '64' in code for better code readability.

Changes in v3:
- No change.

Changes in v2:
- No change

 drivers/gpio/gpio-xilinx.c | 64 +++++++++++++++++++-------------------
 1 file changed, 32 insertions(+), 32 deletions(-)
drivers/gpio/gpio-xilinx.c +32 -32
index 67f9f82e0db0..f86bee271246 100644
@@ -138,37 +138,37 @@ static void xgpio_set_multiple(struct gpio_chip *gc, unsigned long *mask,
 {
 	unsigned long flags;
 	struct xgpio_instance *chip = gpiochip_get_data(gc);
-	int index = xgpio_index(chip, 0);
-	int offset, i;
-
-	spin_lock_irqsave(&chip->gpio_lock[index], flags);
-
-	/* Write to GPIO signals */
-	for (i = 0; i < gc->ngpio; i++) {
-		if (*mask == 0)
-			break;
-		/* Once finished with an index write it out to the register */
-		if (index !=  xgpio_index(chip, i)) {
-			xgpio_writereg(chip->regs + XGPIO_DATA_OFFSET +
-				       index * XGPIO_CHANNEL_OFFSET,
-				       chip->gpio_state[index]);
-			spin_unlock_irqrestore(&chip->gpio_lock[index], flags);
-			index =  xgpio_index(chip, i);
-			spin_lock_irqsave(&chip->gpio_lock[index], flags);
-		}
-		if (__test_and_clear_bit(i, mask)) {
-			offset =  xgpio_offset(chip, i);
-			if (test_bit(i, bits))
-				chip->gpio_state[index] |= BIT(offset);
-			else
-				chip->gpio_state[index] &= ~BIT(offset);
-		}
-	}
-
-	xgpio_writereg(chip->regs + XGPIO_DATA_OFFSET +
-		       index * XGPIO_CHANNEL_OFFSET, chip->gpio_state[index]);
-
-	spin_unlock_irqrestore(&chip->gpio_lock[index], flags);
+	u32 *const state = chip->gpio_state;
+	unsigned int *const width = chip->gpio_width;
+
+	DECLARE_BITMAP(old, 64);
+	DECLARE_BITMAP(new, 64);
+	DECLARE_BITMAP(changed, 64);
+
+	spin_lock_irqsave(&chip->gpio_lock[0], flags);
+	spin_lock(&chip->gpio_lock[1]);
+
+	bitmap_set_value(old, state[0], 0, width[0]);
+	bitmap_set_value(old, state[1], width[0], width[1]);
+	bitmap_replace(new, old, bits, mask, gc->ngpio);
+
+	bitmap_set_value(old, state[0], 0, 32);
+	bitmap_set_value(old, state[1], 32, 32);
+	state[0] = bitmap_get_value(new, 0, width[0]);
+	state[1] = bitmap_get_value(new, width[0], width[1]);
+	bitmap_set_value(new, state[0], 0, 32);
+	bitmap_set_value(new, state[1], 32, 32);
+	bitmap_xor(changed, old, new, 64);
+
+	if (((u32 *)changed)[0])
+		xgpio_writereg(chip->regs + XGPIO_DATA_OFFSET,
+				state[0]);
+	if (((u32 *)changed)[1])
+		xgpio_writereg(chip->regs + XGPIO_DATA_OFFSET +
+				XGPIO_CHANNEL_OFFSET, state[1]);
+
+	spin_unlock(&chip->gpio_lock[1]);
+	spin_unlock_irqrestore(&chip->gpio_lock[0], flags);
 }
 
 /**
@@ -292,6 +292,7 @@ static int xgpio_probe(struct platform_device *pdev)
 		chip->gpio_width[0] = 32;
 
 	spin_lock_init(&chip->gpio_lock[0]);
+	spin_lock_init(&chip->gpio_lock[1]);
 
 	if (of_property_read_u32(np, "xlnx,is-dual", &is_dual))
 		is_dual = 0;
@@ -314,7 +315,6 @@ static int xgpio_probe(struct platform_device *pdev)
 					 &chip->gpio_width[1]))
 			chip->gpio_width[1] = 32;
 
-		spin_lock_init(&chip->gpio_lock[1]);
 	}
 
 	chip->gc.base = -1;
Signature · 1 line
2.26.2
#10 lore raw
In reply to Syed Nayyar Waris
Quoted 2 lines · Syed Nayyar Waris

Since this patchset primarily affects GPIO drivers, would you like to pick it up through your GPIO tree?

Definitely will, once we are finished!

I see Andy still has comments and we need more iterations. That is fine, because we are not in any hurry. Just keep posting it!

Let's merge this for v5.11 when we are finished with it.

Yours,
Linus Walleij

#11 lore raw
Quoted 8 lines · Linus Walleij
Quoted 2 lines · Syed Nayyar Waris

Since this patchset primarily affects GPIO drivers, would you like to pick it up through your GPIO tree?

Definitely will, once we are finished!

Quoted 8 lines

I see Andy still has comments and we need more iterations. That is fine, because we are not in any hurry. Just keep posting it!

Let's merge this for v5.11 when we are finished with it.

Yours,
Linus Walleij

Hi Linus,

Just thought of giving an update. The V-12 revision of the patchset has been submitted (19 Oct).

Let me know if you face any questions regarding the V-12 patchset. Thanks !

Regards
Syed Nayyar Waris