2020-08-13 07:06:01

by Wright Feng

[permalink] [raw]
Subject: [PATCH] brcmfmac: reserve tx credit only when txctl is ready to send

The 4329 throughput drops from 40.2 Mbits/sec to 190 Kbits/sec in 2G
11n mode because the commit b41c232d33666 ("brcmfmac: reserve 2 credits
for host tx control path"). To fix the issue, host driver only reserves
tx control credit when there is a txctl frame is pending to send. And
we also check available credit by using "not equal to 0" instead of
"greater than 0" because tx_max and tx_seq are circled positive numbers.

Reported-by: Dmitry Osipenko <[email protected]>
Fixes: b41c232d33666 ("brcmfmac: reserve 2 credits for host tx control
path")
Signed-off-by: Wright Feng <[email protected]>
Tested-by: Dmitry Osipenko <[email protected]>
---
.../net/wireless/broadcom/brcm80211/brcmfmac/sdio.c | 12 +++++++++---
1 file changed, 9 insertions(+), 3 deletions(-)

diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/sdio.c b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/sdio.c
index e8712ad3ac45..3c07d1bbe1c6 100644
--- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/sdio.c
+++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/sdio.c
@@ -664,9 +664,15 @@ static void pkt_align(struct sk_buff *p, int len, int align)
/* To check if there's window offered */
static bool data_ok(struct brcmf_sdio *bus)
{
- /* Reserve TXCTL_CREDITS credits for txctl */
- return (bus->tx_max - bus->tx_seq) > TXCTL_CREDITS &&
- ((bus->tx_max - bus->tx_seq) & 0x80) == 0;
+ u8 tx_rsv = 0;
+
+ /* Reserve TXCTL_CREDITS credits for txctl when it is ready to send */
+ if (bus->ctrl_frame_stat)
+ tx_rsv = TXCTL_CREDITS;
+
+ return (bus->tx_max - bus->tx_seq - tx_rsv) != 0 &&
+ ((bus->tx_max - bus->tx_seq - tx_rsv) & 0x80) == 0;
+
}

/* To check if there's window offered */
--
2.25.0


2020-08-13 11:02:25

by Kalle Valo

[permalink] [raw]
Subject: Re: [PATCH] brcmfmac: reserve tx credit only when txctl is ready to send

Wright Feng <[email protected]> writes:

> The 4329 throughput drops from 40.2 Mbits/sec to 190 Kbits/sec in 2G
> 11n mode because the commit b41c232d33666 ("brcmfmac: reserve 2 credits
> for host tx control path"). To fix the issue, host driver only reserves
> tx control credit when there is a txctl frame is pending to send. And
> we also check available credit by using "not equal to 0" instead of
> "greater than 0" because tx_max and tx_seq are circled positive numbers.
>
> Reported-by: Dmitry Osipenko <[email protected]>
> Fixes: b41c232d33666 ("brcmfmac: reserve 2 credits for host tx control
> path")

Fixes tag should be in one line, but I can fix it.

> Signed-off-by: Wright Feng <[email protected]>
> Tested-by: Dmitry Osipenko <[email protected]>

And I'll queue this to v5.9.

--
https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

2020-08-18 14:37:11

by Kalle Valo

[permalink] [raw]
Subject: Re: [PATCH] brcmfmac: reserve tx credit only when txctl is ready to send

Wright Feng <[email protected]> wrote:

> The 4329 throughput drops from 40.2 Mbits/sec to 190 Kbits/sec in 2G
> 11n mode because the commit b41c232d33666 ("brcmfmac: reserve 2 credits
> for host tx control path"). To fix the issue, host driver only reserves
> tx control credit when there is a txctl frame is pending to send. And
> we also check available credit by using "not equal to 0" instead of
> "greater than 0" because tx_max and tx_seq are circled positive numbers.
>
> Reported-by: Dmitry Osipenko <[email protected]>
> Fixes: b41c232d33666 ("brcmfmac: reserve 2 credits for host tx control path")
> Signed-off-by: Wright Feng <[email protected]>
> Tested-by: Dmitry Osipenko <[email protected]>

Patch applied to wireless-drivers.git, thanks.

9a2a0862d973 brcmfmac: reserve tx credit only when txctl is ready to send

--
https://patchwork.kernel.org/patch/11711809/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches