[PATCH v2] arm64: Implement optimised IP checksum helpers

[PATCH v2] arm64: Implement optimised IP checksum helpers

[Robin Murphy]

AArch64 is capable of 128-bit memory accesses without alignment
restrictions, which makes it both possible and highly practical to slurp
up a typical 20-byte IP header in just 2 loads. Implement our own
version of ip_fast_checksum() to take advantage of that, resulting in
considerably fewer instructions and memory accesses than the generic
version. We can also get more optimal code generation for csum_fold() by
defining it a slightly different way round from the generic version, so
throw that into the mix too.

Suggested-by: Luke Starrett <luke.starrett@broadcom.com>
Acked-by: Luke Starrett <luke.starrett@broadcom.com>
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
---

Minor changes: include types.h for correctness, add Luke's ack.

 arch/arm64/include/asm/checksum.h | 51 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 51 insertions(+)
 create mode 100644 arch/arm64/include/asm/checksum.h

diff --git a/arch/arm64/include/asm/checksum.h b/arch/arm64/include/asm/checksum.h
new file mode 100644
index 000000000000..09f65339d66d
--- /dev/null
+++ b/arch/arm64/include/asm/checksum.h
@@ -0,0 +1,51 @@
+/*
+ * Copyright (C) 2016 ARM Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+#ifndef __ASM_CHECKSUM_H
+#define __ASM_CHECKSUM_H
+
+#include <linux/types.h>
+
+static inline __sum16 csum_fold(__wsum csum)
+{
+	u32 sum = (__force u32)csum;
+	sum += (sum >> 16) | (sum << 16);
+	return ~(__force __sum16)(sum >> 16);
+}
+#define csum_fold csum_fold
+
+static inline __sum16 ip_fast_csum(const void *iph, unsigned int ihl)
+{
+	__uint128_t tmp;
+	u64 sum;
+
+	tmp = *(const __uint128_t *)iph;
+	iph += 16;
+	ihl -= 4;
+	tmp += ((tmp >> 64) | (tmp << 64));
+	sum = tmp >> 64;
+	do {
+		sum += *(const u32 *)iph;
+		iph += 4;
+	} while (--ihl);
+
+	sum += ((sum >> 32) | (sum << 32));
+	return csum_fold(sum >> 32);
+}
+#define ip_fast_csum ip_fast_csum
+
+#include <asm-generic/checksum.h>
+
+#endif	/* __ASM_CHECKSUM_H */
-- 
2.8.1.dirty

[PATCH v2] arm64: Implement optimised IP checksum helpers

[James Morse]

Hi Robin,

On 31/05/16 12:22, Robin Murphy wrote:
> AArch64 is capable of 128-bit memory accesses without alignment
> restrictions, which makes it both possible and highly practical to slurp
> up a typical 20-byte IP header in just 2 loads. Implement our own
> version of ip_fast_checksum() to take advantage of that, resulting in
> considerably fewer instructions and memory accesses than the generic
> version. We can also get more optimal code generation for csum_fold() by
> defining it a slightly different way round from the generic version, so
> throw that into the mix too.
> 
> Suggested-by: Luke Starrett <luke.starrett@broadcom.com>
> Acked-by: Luke Starrett <luke.starrett@broadcom.com>
> Signed-off-by: Robin Murphy <robin.murphy@arm.com>
> ---
> 
> Minor changes: include types.h for correctness, add Luke's ack.
> 
>  arch/arm64/include/asm/checksum.h | 51 +++++++++++++++++++++++++++++++++++++++

Maybe a nit, don't you need to remove the 'generic-y += checksum.h' line from
arch/arm64/include/asm/Kbuild to avoid the generated version being created too? [0]

The compiler on my box picks your header in preference to the generated one, but
[1] suggests it isn't to be trusted!


Thanks,

James

[0] d8ecc5cd8e22 ("kbuild: asm-generic support")
[1] https://lkml.org/lkml/2016/5/23/78

[PATCH v2] arm64: Implement optimised IP checksum helpers

[Robin Murphy]

Hi James,

On 31/05/16 15:24, James Morse wrote:
> Hi Robin,
>
> On 31/05/16 12:22, Robin Murphy wrote:
>> AArch64 is capable of 128-bit memory accesses without alignment
>> restrictions, which makes it both possible and highly practical to slurp
>> up a typical 20-byte IP header in just 2 loads. Implement our own
>> version of ip_fast_checksum() to take advantage of that, resulting in
>> considerably fewer instructions and memory accesses than the generic
>> version. We can also get more optimal code generation for csum_fold() by
>> defining it a slightly different way round from the generic version, so
>> throw that into the mix too.
>>
>> Suggested-by: Luke Starrett <luke.starrett@broadcom.com>
>> Acked-by: Luke Starrett <luke.starrett@broadcom.com>
>> Signed-off-by: Robin Murphy <robin.murphy@arm.com>
>> ---
>>
>> Minor changes: include types.h for correctness, add Luke's ack.
>>
>>   arch/arm64/include/asm/checksum.h | 51 +++++++++++++++++++++++++++++++++++++++
>
> Maybe a nit, don't you need to remove the 'generic-y += checksum.h' line from
> arch/arm64/include/asm/Kbuild to avoid the generated version being created too? [0]

I guess you skimmed the very end of the patch ;)

+#include <asm-generic/checksum.h>

> The compiler on my box picks your header in preference to the generated one, but
> [1] suggests it isn't to be trusted!

We only cherry-pick a couple of functions to optimise, so not building 
the generic file would necessitate copy-pasting the rest of it. I'd hope 
this pattern is sufficiently robust, since it's already in use on a 
number of other architectures! AFAICS [1] only applies when the generic 
file shouldn't be built at all, so cross fingers we're probably OK.

Robin.

> Thanks,
>
> James
>
> [0] d8ecc5cd8e22 ("kbuild: asm-generic support")
> [1] https://lkml.org/lkml/2016/5/23/78
>
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel at lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
>

[PATCH v2] arm64: Implement optimised IP checksum helpers

[Robin Murphy]

On 31/05/16 15:39, Robin Murphy wrote:
> Hi James,
>
> On 31/05/16 15:24, James Morse wrote:
>> Hi Robin,
>>
>> On 31/05/16 12:22, Robin Murphy wrote:
>>> AArch64 is capable of 128-bit memory accesses without alignment
>>> restrictions, which makes it both possible and highly practical to slurp
>>> up a typical 20-byte IP header in just 2 loads. Implement our own
>>> version of ip_fast_checksum() to take advantage of that, resulting in
>>> considerably fewer instructions and memory accesses than the generic
>>> version. We can also get more optimal code generation for csum_fold() by
>>> defining it a slightly different way round from the generic version, so
>>> throw that into the mix too.
>>>
>>> Suggested-by: Luke Starrett <luke.starrett@broadcom.com>
>>> Acked-by: Luke Starrett <luke.starrett@broadcom.com>
>>> Signed-off-by: Robin Murphy <robin.murphy@arm.com>
>>> ---
>>>
>>> Minor changes: include types.h for correctness, add Luke's ack.
>>>
>>>   arch/arm64/include/asm/checksum.h | 51
>>> +++++++++++++++++++++++++++++++++++++++
>>
>> Maybe a nit, don't you need to remove the 'generic-y += checksum.h'
>> line from
>> arch/arm64/include/asm/Kbuild to avoid the generated version being
>> created too? [0]
>
> I guess you skimmed the very end of the patch ;)
>
> +#include <asm-generic/checksum.h>

Oh wait, I'm being a massive idiot - include/asm-generic is not the same 
thing as arch/arm/include/generated/asm, is it? I see what you mean now, 
and yes it probably does want fixing...

Sorry!
Robin.

>> The compiler on my box picks your header in preference to the
>> generated one, but
>> [1] suggests it isn't to be trusted!
>
> We only cherry-pick a couple of functions to optimise, so not building
> the generic file would necessitate copy-pasting the rest of it. I'd hope
> this pattern is sufficiently robust, since it's already in use on a
> number of other architectures! AFAICS [1] only applies when the generic
> file shouldn't be built at all, so cross fingers we're probably OK.
>
> Robin.
>
>> Thanks,
>>
>> James
>>
>> [0] d8ecc5cd8e22 ("kbuild: asm-generic support")
>> [1] https://lkml.org/lkml/2016/5/23/78
>>
>> _______________________________________________
>> linux-arm-kernel mailing list
>> linux-arm-kernel at lists.infradead.org
>> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
>>
>
>
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel at lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v3] arm64: Implement optimised IP checksum helpers

[Robin Murphy]

AArch64 is capable of 128-bit memory accesses without alignment
restrictions, which makes it both possible and highly practical to slurp
up a typical 20-byte IP header in just 2 loads. Implement our own
version of ip_fast_checksum() to take advantage of that, resulting in
considerably fewer instructions and memory accesses than the generic
version. We can also get more optimal code generation for csum_fold() by
defining it a slightly different way round from the generic version, so
throw that into the mix too.

Suggested-by: Luke Starrett <luke.starrett@broadcom.com>
Acked-by: Luke Starrett <luke.starrett@broadcom.com>
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
---

v3: Don't generate generic header [James]
v2: Include types.h, add Luke's ack

 arch/arm64/include/asm/Kbuild     |  1 -
 arch/arm64/include/asm/checksum.h | 51 +++++++++++++++++++++++++++++++++++++++
 2 files changed, 51 insertions(+), 1 deletion(-)
 create mode 100644 arch/arm64/include/asm/checksum.h

diff --git a/arch/arm64/include/asm/Kbuild b/arch/arm64/include/asm/Kbuild
index cff532a6744e..f43d2c44c765 100644
--- a/arch/arm64/include/asm/Kbuild
+++ b/arch/arm64/include/asm/Kbuild
@@ -1,6 +1,5 @@
 generic-y += bug.h
 generic-y += bugs.h
-generic-y += checksum.h
 generic-y += clkdev.h
 generic-y += cputime.h
 generic-y += current.h
diff --git a/arch/arm64/include/asm/checksum.h b/arch/arm64/include/asm/checksum.h
new file mode 100644
index 000000000000..09f65339d66d
--- /dev/null
+++ b/arch/arm64/include/asm/checksum.h
@@ -0,0 +1,51 @@
+/*
+ * Copyright (C) 2016 ARM Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+#ifndef __ASM_CHECKSUM_H
+#define __ASM_CHECKSUM_H
+
+#include <linux/types.h>
+
+static inline __sum16 csum_fold(__wsum csum)
+{
+	u32 sum = (__force u32)csum;
+	sum += (sum >> 16) | (sum << 16);
+	return ~(__force __sum16)(sum >> 16);
+}
+#define csum_fold csum_fold
+
+static inline __sum16 ip_fast_csum(const void *iph, unsigned int ihl)
+{
+	__uint128_t tmp;
+	u64 sum;
+
+	tmp = *(const __uint128_t *)iph;
+	iph += 16;
+	ihl -= 4;
+	tmp += ((tmp >> 64) | (tmp << 64));
+	sum = tmp >> 64;
+	do {
+		sum += *(const u32 *)iph;
+		iph += 4;
+	} while (--ihl);
+
+	sum += ((sum >> 32) | (sum << 32));
+	return csum_fold(sum >> 32);
+}
+#define ip_fast_csum ip_fast_csum
+
+#include <asm-generic/checksum.h>
+
+#endif	/* __ASM_CHECKSUM_H */
-- 
2.8.1.dirty

[PATCH v3] arm64: Implement optimised IP checksum helpers

[Goel, Sameer]

Thanks for the arm64 checksum file. I saw a 4 fold speedup when 
calculating checksum for a 20 byte buffer on my test platform.

Thanks,
Sameer

On 5/31/2016 11:04 AM, Robin Murphy wrote:
> AArch64 is capable of 128-bit memory accesses without alignment
> restrictions, which makes it both possible and highly practical to slurp
> up a typical 20-byte IP header in just 2 loads. Implement our own
> version of ip_fast_checksum() to take advantage of that, resulting in
> considerably fewer instructions and memory accesses than the generic
> version. We can also get more optimal code generation for csum_fold() by
> defining it a slightly different way round from the generic version, so
> throw that into the mix too.
>
> Suggested-by: Luke Starrett <luke.starrett@broadcom.com>
> Acked-by: Luke Starrett <luke.starrett@broadcom.com>
> Signed-off-by: Robin Murphy <robin.murphy@arm.com>
> ---
>
> v3: Don't generate generic header [James]
> v2: Include types.h, add Luke's ack
>
>  arch/arm64/include/asm/Kbuild     |  1 -
>  arch/arm64/include/asm/checksum.h | 51 +++++++++++++++++++++++++++++++++++++++
>  2 files changed, 51 insertions(+), 1 deletion(-)
>  create mode 100644 arch/arm64/include/asm/checksum.h
>
> diff --git a/arch/arm64/include/asm/Kbuild b/arch/arm64/include/asm/Kbuild
> index cff532a6744e..f43d2c44c765 100644
> --- a/arch/arm64/include/asm/Kbuild
> +++ b/arch/arm64/include/asm/Kbuild
> @@ -1,6 +1,5 @@
>  generic-y += bug.h
>  generic-y += bugs.h
> -generic-y += checksum.h
>  generic-y += clkdev.h
>  generic-y += cputime.h
>  generic-y += current.h
> diff --git a/arch/arm64/include/asm/checksum.h b/arch/arm64/include/asm/checksum.h
> new file mode 100644
> index 000000000000..09f65339d66d
> --- /dev/null
> +++ b/arch/arm64/include/asm/checksum.h
> @@ -0,0 +1,51 @@
> +/*
> + * Copyright (C) 2016 ARM Ltd.
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program.  If not, see <http://www.gnu.org/licenses/>.
> + */
> +#ifndef __ASM_CHECKSUM_H
> +#define __ASM_CHECKSUM_H
> +
> +#include <linux/types.h>
> +
> +static inline __sum16 csum_fold(__wsum csum)
> +{
> +	u32 sum = (__force u32)csum;
> +	sum += (sum >> 16) | (sum << 16);
> +	return ~(__force __sum16)(sum >> 16);
> +}
> +#define csum_fold csum_fold
> +
> +static inline __sum16 ip_fast_csum(const void *iph, unsigned int ihl)
> +{
> +	__uint128_t tmp;
> +	u64 sum;
> +
> +	tmp = *(const __uint128_t *)iph;
> +	iph += 16;
> +	ihl -= 4;
> +	tmp += ((tmp >> 64) | (tmp << 64));
> +	sum = tmp >> 64;
> +	do {
> +		sum += *(const u32 *)iph;
> +		iph += 4;
> +	} while (--ihl);
> +
> +	sum += ((sum >> 32) | (sum << 32));
> +	return csum_fold(sum >> 32);
> +}
> +#define ip_fast_csum ip_fast_csum
> +
> +#include <asm-generic/checksum.h>
> +
> +#endif	/* __ASM_CHECKSUM_H */
>

-- 
Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project.

[PATCH v3] arm64: Implement optimised IP checksum helpers

[Catalin Marinas]

On Tue, May 31, 2016 at 06:04:40PM +0100, Robin Murphy wrote:
> AArch64 is capable of 128-bit memory accesses without alignment
> restrictions, which makes it both possible and highly practical to slurp
> up a typical 20-byte IP header in just 2 loads. Implement our own
> version of ip_fast_checksum() to take advantage of that, resulting in
> considerably fewer instructions and memory accesses than the generic
> version. We can also get more optimal code generation for csum_fold() by
> defining it a slightly different way round from the generic version, so
> throw that into the mix too.
> 
> Suggested-by: Luke Starrett <luke.starrett@broadcom.com>
> Acked-by: Luke Starrett <luke.starrett@broadcom.com>
> Signed-off-by: Robin Murphy <robin.murphy@arm.com>

I now applied the correct version. Thanks for pointing out.

-- 
Catalin