All of lore.kernel.org
 help / color / mirror / Atom feed
From: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
To: Giuseppe CAVALLARO <peppe.cavallaro@st.com>
Cc: netdev@vger.kernel.org, linux-amlogic@lists.infradead.org,
	Alexandre Torgue <alexandre.torgue@st.com>,
	Johnson Leung <r58129@freescale.com>
Subject: Re: stmmac/RTL8211F/Meson GXBB: TX throughput problems
Date: Sat, 1 Oct 2016 22:40:59 +0200	[thread overview]
Message-ID: <CAFBinCAA9J4o2cs4ruSuPRH4qK6izDoxJCneE=Uogk0c+_DSkQ@mail.gmail.com> (raw)
In-Reply-To: <CAFBinCC0HiEXx5mELztH6jYfUq2f1u0Tb+pSzAvwQSu9vYydGQ@mail.gmail.com>

[-- Attachment #1: Type: text/plain, Size: 1038 bytes --]

On Sat, Oct 1, 2016 at 10:15 PM, Martin Blumenstingl
<martin.blumenstingl@googlemail.com> wrote:
> On Wed, Sep 14, 2016 at 5:30 PM, Giuseppe CAVALLARO
> <peppe.cavallaro@st.com> wrote:
>> Also you could check the AXI tuning and PBL value. To be honest
>> (thinking about your problem) I can actually suspect some related
>> problem on bus setup. So I suggest you to play with these value
>> (better if you ask for having values from HW validation on your side).
>> Otherwise the stmmac uses a default that cannot be good for your
>> platform. For example, sometime I have seen that PBL is better if
>> reduced to 8 instead of 32 and w/o 4xPBL...
> unfortunately the original kernel/firmware shipped with my device
> seems to be also affected by that TX bug, so comparing with the
> "stock" values does not work for me.
I tried the following changes:
- snps,pbl = <4>; in .dts
- snps,pbl = <2>; in .dts
- snps,pbl = <8>; in .dts and disabled DMA_BUS_MODE_MAXPBL (= 4xPBL) in stmmac

None of these seem to have any impact on the TX problem

[-- Attachment #2: meson-stmmac-pbl.patch --]
[-- Type: text/x-patch, Size: 1012 bytes --]

diff --git a/arch/arm64/boot/dts/amlogic/meson-gxbb.dtsi b/arch/arm64/boot/dts/amlogic/meson-gxbb.dtsi
index 2e89ec4..7deab7a 100644
--- a/arch/arm64/boot/dts/amlogic/meson-gxbb.dtsi
+++ b/arch/arm64/boot/dts/amlogic/meson-gxbb.dtsi
@@ -698,6 +698,7 @@
 				 <&clkc CLKID_MPLL2>;
 			clock-names = "stmmaceth", "clkin0", "clkin1";
 			phy-mode = "rgmii";
+			snps,pbl = <8>;
 			status = "disabled";
 		};
 	};
diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac1000_dma.c b/drivers/net/ethernet/stmicro/stmmac/dwmac1000_dma.c
index 99074695..fffb407 100644
--- a/drivers/net/ethernet/stmicro/stmmac/dwmac1000_dma.c
+++ b/drivers/net/ethernet/stmicro/stmmac/dwmac1000_dma.c
@@ -97,7 +97,7 @@ static void dwmac1000_dma_init(void __iomem *ioaddr, int pbl, int fb, int mb,
 	 * so only the bits: 13-8 are programmed with the PBL passed from the
 	 * platform.
 	 */
-	value |= DMA_BUS_MODE_MAXPBL;
+	value &= ~DMA_BUS_MODE_MAXPBL;
 	value &= ~DMA_BUS_MODE_PBL_MASK;
 	value |= (pbl << DMA_BUS_MODE_PBL_SHIFT);
 

WARNING: multiple messages have this Message-ID (diff)
From: martin.blumenstingl@googlemail.com (Martin Blumenstingl)
To: linus-amlogic@lists.infradead.org
Subject: stmmac/RTL8211F/Meson GXBB: TX throughput problems
Date: Sat, 1 Oct 2016 22:40:59 +0200	[thread overview]
Message-ID: <CAFBinCAA9J4o2cs4ruSuPRH4qK6izDoxJCneE=Uogk0c+_DSkQ@mail.gmail.com> (raw)
In-Reply-To: <CAFBinCC0HiEXx5mELztH6jYfUq2f1u0Tb+pSzAvwQSu9vYydGQ@mail.gmail.com>

On Sat, Oct 1, 2016 at 10:15 PM, Martin Blumenstingl
<martin.blumenstingl@googlemail.com> wrote:
> On Wed, Sep 14, 2016 at 5:30 PM, Giuseppe CAVALLARO
> <peppe.cavallaro@st.com> wrote:
>> Also you could check the AXI tuning and PBL value. To be honest
>> (thinking about your problem) I can actually suspect some related
>> problem on bus setup. So I suggest you to play with these value
>> (better if you ask for having values from HW validation on your side).
>> Otherwise the stmmac uses a default that cannot be good for your
>> platform. For example, sometime I have seen that PBL is better if
>> reduced to 8 instead of 32 and w/o 4xPBL...
> unfortunately the original kernel/firmware shipped with my device
> seems to be also affected by that TX bug, so comparing with the
> "stock" values does not work for me.
I tried the following changes:
- snps,pbl = <4>; in .dts
- snps,pbl = <2>; in .dts
- snps,pbl = <8>; in .dts and disabled DMA_BUS_MODE_MAXPBL (= 4xPBL) in stmmac

None of these seem to have any impact on the TX problem
-------------- next part --------------
A non-text attachment was scrubbed...
Name: meson-stmmac-pbl.patch
Type: text/x-patch
Size: 1012 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-amlogic/attachments/20161001/8e44374c/attachment.bin>

  reply	other threads:[~2016-10-01 20:41 UTC|newest]

Thread overview: 56+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-09-11 20:39 stmmac/RTL8211F/Meson GXBB: TX throughput problems Martin Blumenstingl
2016-09-11 20:39 ` Martin Blumenstingl
2016-09-12 16:37 ` Alexandre Torgue
2016-09-12 16:37   ` Alexandre Torgue
2016-09-12 21:26   ` Martin Blumenstingl
2016-09-12 21:26     ` Martin Blumenstingl
2016-09-17 21:23     ` André Roth
2016-09-17 21:23       ` André Roth
2016-09-26  6:17       ` Giuseppe CAVALLARO
2016-09-26  6:17         ` Giuseppe CAVALLARO
2016-10-01 15:58         ` Martin Blumenstingl
2016-10-01 15:58           ` Martin Blumenstingl
2016-11-03 16:36           ` Jerome Brunet
2016-11-03 16:36             ` Jerome Brunet
2016-11-05 12:20             ` Martin Blumenstingl
2016-11-05 12:20               ` Martin Blumenstingl
2016-11-13 19:20             ` André Roth
2016-11-13 19:20               ` André Roth
     [not found] ` <216F2694-1C1D-44DA-AC15-57ED15C24BBE@bluematt.me>
2016-09-12 21:29   ` Martin Blumenstingl
2016-09-12 21:29     ` Martin Blumenstingl
2016-09-14 15:30 ` Giuseppe CAVALLARO
2016-09-14 15:30   ` Giuseppe CAVALLARO
2016-09-18 20:42   ` André Roth
2016-09-18 20:42     ` André Roth
2016-09-19  5:38     ` Giuseppe CAVALLARO
2016-09-19  5:38       ` Giuseppe CAVALLARO
2016-10-01 20:15   ` Martin Blumenstingl
2016-10-01 20:15     ` Martin Blumenstingl
2016-10-01 20:40     ` Martin Blumenstingl [this message]
2016-10-01 20:40       ` Martin Blumenstingl
2016-10-31 10:25     ` André Roth
2016-10-31 10:25       ` André Roth
2016-11-03 16:57       ` Jerome Brunet
2016-11-03 16:57         ` Jerome Brunet
2016-11-07 10:59         ` Giuseppe CAVALLARO
2016-11-07 10:59           ` Giuseppe CAVALLARO
2016-11-07 17:37           ` Martin Blumenstingl
2016-11-07 17:37             ` Martin Blumenstingl
2016-11-14  7:47             ` Giuseppe CAVALLARO
2016-11-14  7:47               ` Giuseppe CAVALLARO
2016-11-14 11:02               ` Jerome Brunet
2016-11-14 11:02                 ` Jerome Brunet
2016-11-14 15:00             ` Andreas Färber
2016-11-14 15:00               ` Andreas Färber
2016-11-15 11:27               ` Jerome Brunet
2016-11-15 11:27                 ` Jerome Brunet
2016-11-13 19:13         ` André Roth
2016-11-13 19:13           ` André Roth
2016-11-14 10:49           ` Jerome Brunet
2016-11-14 10:49             ` Jerome Brunet
2016-11-17 18:44             ` André Roth
2016-11-17 18:44               ` André Roth
2016-11-17 21:47               ` Jerome Brunet
2016-11-17 21:47                 ` Jerome Brunet
2016-10-01 20:26   ` Martin Blumenstingl
2016-10-01 20:26     ` Martin Blumenstingl

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CAFBinCAA9J4o2cs4ruSuPRH4qK6izDoxJCneE=Uogk0c+_DSkQ@mail.gmail.com' \
    --to=martin.blumenstingl@googlemail.com \
    --cc=alexandre.torgue@st.com \
    --cc=linux-amlogic@lists.infradead.org \
    --cc=netdev@vger.kernel.org \
    --cc=peppe.cavallaro@st.com \
    --cc=r58129@freescale.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.