bcm53xx: revert bgmac back to the old limited max frame size

Bumping max frame size has significantly affected network performance.
It was done by upstream commit that first appeared in the 5.7 release.

This change bumps NAT masquarade speed from 196 Mb/s to 383 Mb/s for the
BCM4708 SoC.

Ref: f55f1dbaad ("bcm53xx: switch to the kernel 5.10")
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
This commit is contained in:
Rafał Miłecki 2022-06-10 14:02:19 +02:00
parent d75bb744ea
commit 230c9da963

View File

@ -0,0 +1,33 @@
From: =?UTF-8?q?Rafa=C5=82=20Mi=C5=82ecki?= <rafal@milecki.pl>
Date: Fri, 10 Jun 2022 13:10:47 +0200
Subject: [PATCH] bgmac: reduce max frame size to support just MTU 1500
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
bgmac allocates new replacement buffer before handling each received
frame. Allocating & DMA-preparing 9724 B each time consumes a lot of CPU
time. Ideally bgmac should just respect currently set MTU but it isn't
the case right now. For now just revert back to the old limited frame
size.
This change bumps NAT masquarade speed by ~95%.
Ref: 8c7da63978f1 ("bgmac: configure MTU and add support for frames beyond 8192 byte size")
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
---
drivers/net/ethernet/broadcom/bgmac.h | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
--- a/drivers/net/ethernet/broadcom/bgmac.h
+++ b/drivers/net/ethernet/broadcom/bgmac.h
@@ -366,8 +366,7 @@
#define BGMAC_RX_FRAME_OFFSET 30 /* There are 2 unused bytes between header and real data */
#define BGMAC_RX_BUF_OFFSET (NET_SKB_PAD + NET_IP_ALIGN - \
BGMAC_RX_FRAME_OFFSET)
-/* Jumbo frame size with FCS */
-#define BGMAC_RX_MAX_FRAME_SIZE 9724
+#define BGMAC_RX_MAX_FRAME_SIZE 1536
#define BGMAC_RX_BUF_SIZE (BGMAC_RX_FRAME_OFFSET + BGMAC_RX_MAX_FRAME_SIZE)
#define BGMAC_RX_ALLOC_SIZE (SKB_DATA_ALIGN(BGMAC_RX_BUF_SIZE + BGMAC_RX_BUF_OFFSET) + \
SKB_DATA_ALIGN(sizeof(struct skb_shared_info)))