mvebu: enable coherent crypto on armada 37x0

Armada 37xx has coherent bus, similar to Armada 7k/8k. Cache
synchronization consumes a lot of CPU cycles. Enabling coherent DMA
increases IOPS performance up to 4 times. Some numbers:
					Data length
Algo		MB	   16	  64	 128	 256	1024	1424	4096
DES-ECB		1	+21 %	+5 %	+5 %	+7 %	+7 %	+3 %	+20 %
AES-ECB-128	1	+21 %	+6 %	+6 %	+6 %	+9 %	+8 %	+22 %
AES-CBC-128	1	+21 %	+5 %	+5 %	+5 %	+6 %	+7 %	+23 %
AES-CBC-256	1	+23 %	+7 %	+8 %	+6 %	+11 %	+13 %	+20 %

					Data length
Algo		MB	    16	    64	   256	  512	  1024	 1420	 4096	 8192
AES-GCM-128	1	 +44 %	 +42 %	 +31 %	 +32 %	 +27 %	+30 %	+32 %	+30 %
AES-GCM-128	8	+319 %	+326 %	+163 %	+148 %	 +75 %	+72 %	+74 %	+41 %
AES-GCM-128	4096	+123 %	+128 %	 +90 %	 +83 %	+116 %	+59 %	+38 %	+28 %

					Data length
Algo		MB	   16	   64	  256	 1024	 2048	 4096	 8192
MD5		1	+21 %	+15 %	+29 %	+25 %	+50 %	+16 %	+20 %
SHA1		1	+24 %	+22 %	+27 %	+22 %	+18 %	+20 %	+20 %
SHA256		1	+30 %	+24 %	+25 %	+26 %	+21 %	+41 %	+19 %
SHA512		1	 +4 %	 +3 %	 +8 %	+10 %	+24 %	+10 %	+11 %

Tested on Armada 3720. Platform passes testmgr selftests.

Signed-off-by: Aleksander Jan Bajkowski <olek2@wp.pl>
Link: https://github.com/openwrt/openwrt/pull/23646
Signed-off-by: Robert Marko <robimarko@gmail.com>
This commit is contained in:
Aleksander Jan Bajkowski
2026-06-04 22:44:10 +02:00
committed by Robert Marko
parent 32462ea5b2
commit 1480790ea9
2 changed files with 92 additions and 0 deletions
@@ -0,0 +1,46 @@
From 5dd98e0389545b302b0c46cef2264b5b09c01677 Mon Sep 17 00:00:00 2001
From: Aleksander Jan Bajkowski <olek2@wp.pl>
Date: Sun, 24 May 2026 14:44:53 +0200
Subject: [PATCH] arm64: dts: marvell: armada-37xx: mark EIP97 as dma-coherent
Armada 37xx has coherent bus, similar to Armada 7k/8k. Cache
synchronization consumes a lot of CPU cycles. Enabling coherent DMA
increases IOPS performance up to 4 times. Some numbers:
Data length
Algo MB 16 64 128 256 1024 1424 4096
DES-ECB 1 +21 % +5 % +5 % +7 % +7 % +3 % +20 %
AES-ECB-128 1 +21 % +6 % +6 % +6 % +9 % +8 % +22 %
AES-CBC-128 1 +21 % +5 % +5 % +5 % +6 % +7 % +23 %
AES-CBC-256 1 +23 % +7 % +8 % +6 % +11 % +13 % +20 %
Data length
Algo MB 16 64 256 512 1024 1420 4096 8192
AES-GCM-128 1 +44 % +42 % +31 % +32 % +27 % +30 % +32 % +30 %
AES-GCM-128 8 +319 % +326 % +163 % +148 % +75 % +72 % +74 % +41 %
AES-GCM-128 4096 +123 % +128 % +90 % +83 % +116 % +59 % +38 % +28 %
Data length
Algo MB 16 64 256 1024 2048 4096 8192
MD5 1 +21 % +15 % +29 % +25 % +50 % +16 % +20 %
SHA1 1 +24 % +22 % +27 % +22 % +18 % +20 % +20 %
SHA256 1 +30 % +24 % +25 % +26 % +21 % +41 % +19 %
SHA512 1 +4 % +3 % +8 % +10 % +24 % +10 % +11 %
Tested on Armada 3720. Platform passes testmgr selftests.
Signed-off-by: Aleksander Jan Bajkowski <olek2@wp.pl>
Signed-off-by: Gregory CLEMENT <gregory.clement@bootlin.com>
---
arch/arm64/boot/dts/marvell/armada-37xx.dtsi | 1 +
1 file changed, 1 insertion(+)
--- a/arch/arm64/boot/dts/marvell/armada-37xx.dtsi
+++ b/arch/arm64/boot/dts/marvell/armada-37xx.dtsi
@@ -440,6 +440,7 @@
interrupt-names = "ring0", "ring1", "ring2",
"ring3", "eip", "mem";
clocks = <&nb_periph_clk 15>;
+ dma-coherent;
};
rwtm: mailbox@b0000 {
@@ -0,0 +1,46 @@
From 5dd98e0389545b302b0c46cef2264b5b09c01677 Mon Sep 17 00:00:00 2001
From: Aleksander Jan Bajkowski <olek2@wp.pl>
Date: Sun, 24 May 2026 14:44:53 +0200
Subject: [PATCH] arm64: dts: marvell: armada-37xx: mark EIP97 as dma-coherent
Armada 37xx has coherent bus, similar to Armada 7k/8k. Cache
synchronization consumes a lot of CPU cycles. Enabling coherent DMA
increases IOPS performance up to 4 times. Some numbers:
Data length
Algo MB 16 64 128 256 1024 1424 4096
DES-ECB 1 +21 % +5 % +5 % +7 % +7 % +3 % +20 %
AES-ECB-128 1 +21 % +6 % +6 % +6 % +9 % +8 % +22 %
AES-CBC-128 1 +21 % +5 % +5 % +5 % +6 % +7 % +23 %
AES-CBC-256 1 +23 % +7 % +8 % +6 % +11 % +13 % +20 %
Data length
Algo MB 16 64 256 512 1024 1420 4096 8192
AES-GCM-128 1 +44 % +42 % +31 % +32 % +27 % +30 % +32 % +30 %
AES-GCM-128 8 +319 % +326 % +163 % +148 % +75 % +72 % +74 % +41 %
AES-GCM-128 4096 +123 % +128 % +90 % +83 % +116 % +59 % +38 % +28 %
Data length
Algo MB 16 64 256 1024 2048 4096 8192
MD5 1 +21 % +15 % +29 % +25 % +50 % +16 % +20 %
SHA1 1 +24 % +22 % +27 % +22 % +18 % +20 % +20 %
SHA256 1 +30 % +24 % +25 % +26 % +21 % +41 % +19 %
SHA512 1 +4 % +3 % +8 % +10 % +24 % +10 % +11 %
Tested on Armada 3720. Platform passes testmgr selftests.
Signed-off-by: Aleksander Jan Bajkowski <olek2@wp.pl>
Signed-off-by: Gregory CLEMENT <gregory.clement@bootlin.com>
---
arch/arm64/boot/dts/marvell/armada-37xx.dtsi | 1 +
1 file changed, 1 insertion(+)
--- a/arch/arm64/boot/dts/marvell/armada-37xx.dtsi
+++ b/arch/arm64/boot/dts/marvell/armada-37xx.dtsi
@@ -438,6 +438,7 @@
interrupt-names = "ring0", "ring1", "ring2",
"ring3", "eip", "mem";
clocks = <&nb_periph_clk 15>;
+ dma-coherent;
};
rwtm: mailbox@b0000 {