* [PATCH 1/1] net: stmmac: dwmac4: extract MAC_HW_Feature regs MSB first
@ 2023-08-10 9:59 Furong Xu
2023-08-10 12:55 ` Alexandre TORGUE
2023-08-10 13:59 ` Andrew Lunn
0 siblings, 2 replies; 4+ messages in thread
From: Furong Xu @ 2023-08-10 9:59 UTC (permalink / raw)
To: David S. Miller, Alexandre Torgue, Jose Abreu, Eric Dumazet,
Jakub Kicinski, Paolo Abeni, Maxime Coquelin, Joao Pinto,
Simon Horman
Cc: netdev, linux-stm32, linux-arm-kernel, linux-kernel, xfr, rock.xu,
Furong Xu
The bit extraction sequences are a little messy.
No code changes, just reorder them.
Signed-off-by: Furong Xu <0x1207@gmail.com>
---
.../net/ethernet/stmicro/stmmac/dwmac4_dma.c | 54 +++++++++----------
1 file changed, 26 insertions(+), 28 deletions(-)
diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c b/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c
index 84d3a8551b03..3d9e20078193 100644
--- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c
+++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c
@@ -379,27 +379,27 @@ static int dwmac4_get_hw_feature(void __iomem *ioaddr,
{
u32 hw_cap = readl(ioaddr + GMAC_HW_FEATURE0);
- /* MAC HW feature0 */
- dma_cap->mbps_10_100 = (hw_cap & GMAC_HW_FEAT_MIISEL);
- dma_cap->mbps_1000 = (hw_cap & GMAC_HW_FEAT_GMIISEL) >> 1;
- dma_cap->half_duplex = (hw_cap & GMAC_HW_FEAT_HDSEL) >> 2;
- dma_cap->vlhash = (hw_cap & GMAC_HW_FEAT_VLHASH) >> 4;
+ /* MAC HW feature0 */
+ dma_cap->vlins = (hw_cap & GMAC_HW_FEAT_SAVLANINS) >> 27;
dma_cap->multi_addr = (hw_cap & GMAC_HW_FEAT_ADDMAC) >> 18;
- dma_cap->pcs = (hw_cap & GMAC_HW_FEAT_PCSSEL) >> 3;
- dma_cap->sma_mdio = (hw_cap & GMAC_HW_FEAT_SMASEL) >> 5;
- dma_cap->pmt_remote_wake_up = (hw_cap & GMAC_HW_FEAT_RWKSEL) >> 6;
- dma_cap->pmt_magic_frame = (hw_cap & GMAC_HW_FEAT_MGKSEL) >> 7;
- /* MMC */
- dma_cap->rmon = (hw_cap & GMAC_HW_FEAT_MMCSEL) >> 8;
- /* IEEE 1588-2008 */
- dma_cap->atime_stamp = (hw_cap & GMAC_HW_FEAT_TSSEL) >> 12;
- /* 802.3az - Energy-Efficient Ethernet (EEE) */
- dma_cap->eee = (hw_cap & GMAC_HW_FEAT_EEESEL) >> 13;
/* TX and RX csum */
- dma_cap->tx_coe = (hw_cap & GMAC_HW_FEAT_TXCOSEL) >> 14;
dma_cap->rx_coe = (hw_cap & GMAC_HW_FEAT_RXCOESEL) >> 16;
- dma_cap->vlins = (hw_cap & GMAC_HW_FEAT_SAVLANINS) >> 27;
+ dma_cap->tx_coe = (hw_cap & GMAC_HW_FEAT_TXCOSEL) >> 14;
+ /* 802.3az - Energy-Efficient Ethernet (EEE) */
+ dma_cap->eee = (hw_cap & GMAC_HW_FEAT_EEESEL) >> 13;
+ /* IEEE 1588-2008 */
+ dma_cap->atime_stamp = (hw_cap & GMAC_HW_FEAT_TSSEL) >> 12;
dma_cap->arpoffsel = (hw_cap & GMAC_HW_FEAT_ARPOFFSEL) >> 9;
+ /* MMC */
+ dma_cap->rmon = (hw_cap & GMAC_HW_FEAT_MMCSEL) >> 8;
+ dma_cap->pmt_magic_frame = (hw_cap & GMAC_HW_FEAT_MGKSEL) >> 7;
+ dma_cap->pmt_remote_wake_up = (hw_cap & GMAC_HW_FEAT_RWKSEL) >> 6;
+ dma_cap->sma_mdio = (hw_cap & GMAC_HW_FEAT_SMASEL) >> 5;
+ dma_cap->vlhash = (hw_cap & GMAC_HW_FEAT_VLHASH) >> 4;
+ dma_cap->pcs = (hw_cap & GMAC_HW_FEAT_PCSSEL) >> 3;
+ dma_cap->half_duplex = (hw_cap & GMAC_HW_FEAT_HDSEL) >> 2;
+ dma_cap->mbps_1000 = (hw_cap & GMAC_HW_FEAT_GMIISEL) >> 1;
+ dma_cap->mbps_10_100 = (hw_cap & GMAC_HW_FEAT_MIISEL);
/* MAC HW feature1 */
hw_cap = readl(ioaddr + GMAC_HW_FEATURE1);
@@ -408,7 +408,6 @@ static int dwmac4_get_hw_feature(void __iomem *ioaddr,
dma_cap->av = (hw_cap & GMAC_HW_FEAT_AVSEL) >> 20;
dma_cap->tsoen = (hw_cap & GMAC_HW_TSOEN) >> 18;
dma_cap->sphen = (hw_cap & GMAC_HW_FEAT_SPHEN) >> 17;
-
dma_cap->addr64 = (hw_cap & GMAC_HW_ADDR64) >> 14;
switch (dma_cap->addr64) {
case 0:
@@ -424,31 +423,30 @@ static int dwmac4_get_hw_feature(void __iomem *ioaddr,
dma_cap->addr64 = 32;
break;
}
-
/* RX and TX FIFO sizes are encoded as log2(n / 128). Undo that by
* shifting and store the sizes in bytes.
*/
dma_cap->tx_fifo_size = 128 << ((hw_cap & GMAC_HW_TXFIFOSIZE) >> 6);
dma_cap->rx_fifo_size = 128 << ((hw_cap & GMAC_HW_RXFIFOSIZE) >> 0);
+
/* MAC HW feature2 */
hw_cap = readl(ioaddr + GMAC_HW_FEATURE2);
+ /* Number of Auxiliary Snapshot Inputs */
+ dma_cap->aux_snapshot_n = (hw_cap & GMAC_HW_FEAT_AUXSNAPNUM) >> 28;
+ /* PPS output */
+ dma_cap->pps_out_num = (hw_cap & GMAC_HW_FEAT_PPSOUTNUM) >> 24;
/* TX and RX number of channels */
- dma_cap->number_rx_channel =
- ((hw_cap & GMAC_HW_FEAT_RXCHCNT) >> 12) + 1;
dma_cap->number_tx_channel =
((hw_cap & GMAC_HW_FEAT_TXCHCNT) >> 18) + 1;
+ dma_cap->number_rx_channel =
+ ((hw_cap & GMAC_HW_FEAT_RXCHCNT) >> 12) + 1;
/* TX and RX number of queues */
- dma_cap->number_rx_queues =
- ((hw_cap & GMAC_HW_FEAT_RXQCNT) >> 0) + 1;
dma_cap->number_tx_queues =
((hw_cap & GMAC_HW_FEAT_TXQCNT) >> 6) + 1;
- /* PPS output */
- dma_cap->pps_out_num = (hw_cap & GMAC_HW_FEAT_PPSOUTNUM) >> 24;
-
+ dma_cap->number_rx_queues =
+ ((hw_cap & GMAC_HW_FEAT_RXQCNT) >> 0) + 1;
/* IEEE 1588-2002 */
dma_cap->time_stamp = 0;
- /* Number of Auxiliary Snapshot Inputs */
- dma_cap->aux_snapshot_n = (hw_cap & GMAC_HW_FEAT_AUXSNAPNUM) >> 28;
/* MAC HW feature3 */
hw_cap = readl(ioaddr + GMAC_HW_FEATURE3);
--
2.34.1
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH 1/1] net: stmmac: dwmac4: extract MAC_HW_Feature regs MSB first
2023-08-10 9:59 [PATCH 1/1] net: stmmac: dwmac4: extract MAC_HW_Feature regs MSB first Furong Xu
@ 2023-08-10 12:55 ` Alexandre TORGUE
2023-08-10 13:59 ` Andrew Lunn
1 sibling, 0 replies; 4+ messages in thread
From: Alexandre TORGUE @ 2023-08-10 12:55 UTC (permalink / raw)
To: Furong Xu, David S. Miller, Jose Abreu, Eric Dumazet,
Jakub Kicinski, Paolo Abeni, Maxime Coquelin, Joao Pinto,
Simon Horman
Cc: netdev, linux-stm32, linux-arm-kernel, linux-kernel, xfr, rock.xu
Hi Furong
On 8/10/23 11:59, Furong Xu wrote:
> The bit extraction sequences are a little messy.
> No code changes, just reorder them.
>
> Signed-off-by: Furong Xu <0x1207@gmail.com>
> ---
> .../net/ethernet/stmicro/stmmac/dwmac4_dma.c | 54 +++++++++----------
> 1 file changed, 26 insertions(+), 28 deletions(-)
>
> diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c b/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c
> index 84d3a8551b03..3d9e20078193 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c
> +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c
> @@ -379,27 +379,27 @@ static int dwmac4_get_hw_feature(void __iomem *ioaddr,
> {
> u32 hw_cap = readl(ioaddr + GMAC_HW_FEATURE0);
>
> - /* MAC HW feature0 */
Why to not start with the LSB ?
> - dma_cap->mbps_10_100 = (hw_cap & GMAC_HW_FEAT_MIISEL);
> - dma_cap->mbps_1000 = (hw_cap & GMAC_HW_FEAT_GMIISEL) >> 1;
> - dma_cap->half_duplex = (hw_cap & GMAC_HW_FEAT_HDSEL) >> 2;
> - dma_cap->vlhash = (hw_cap & GMAC_HW_FEAT_VLHASH) >> 4;
> + /* MAC HW feature0 */
> + dma_cap->vlins = (hw_cap & GMAC_HW_FEAT_SAVLANINS) >> 27;
> dma_cap->multi_addr = (hw_cap & GMAC_HW_FEAT_ADDMAC) >> 18;
> - dma_cap->pcs = (hw_cap & GMAC_HW_FEAT_PCSSEL) >> 3;
> - dma_cap->sma_mdio = (hw_cap & GMAC_HW_FEAT_SMASEL) >> 5;
> - dma_cap->pmt_remote_wake_up = (hw_cap & GMAC_HW_FEAT_RWKSEL) >> 6;
> - dma_cap->pmt_magic_frame = (hw_cap & GMAC_HW_FEAT_MGKSEL) >> 7;
> - /* MMC */
> - dma_cap->rmon = (hw_cap & GMAC_HW_FEAT_MMCSEL) >> 8;
> - /* IEEE 1588-2008 */
> - dma_cap->atime_stamp = (hw_cap & GMAC_HW_FEAT_TSSEL) >> 12;
> - /* 802.3az - Energy-Efficient Ethernet (EEE) */
> - dma_cap->eee = (hw_cap & GMAC_HW_FEAT_EEESEL) >> 13;
> /* TX and RX csum */
> - dma_cap->tx_coe = (hw_cap & GMAC_HW_FEAT_TXCOSEL) >> 14;
> dma_cap->rx_coe = (hw_cap & GMAC_HW_FEAT_RXCOESEL) >> 16;
> - dma_cap->vlins = (hw_cap & GMAC_HW_FEAT_SAVLANINS) >> 27;
> + dma_cap->tx_coe = (hw_cap & GMAC_HW_FEAT_TXCOSEL) >> 14;
> + /* 802.3az - Energy-Efficient Ethernet (EEE) */
> + dma_cap->eee = (hw_cap & GMAC_HW_FEAT_EEESEL) >> 13;
> + /* IEEE 1588-2008 */
> + dma_cap->atime_stamp = (hw_cap & GMAC_HW_FEAT_TSSEL) >> 12;
> dma_cap->arpoffsel = (hw_cap & GMAC_HW_FEAT_ARPOFFSEL) >> 9;
> + /* MMC */
> + dma_cap->rmon = (hw_cap & GMAC_HW_FEAT_MMCSEL) >> 8;
> + dma_cap->pmt_magic_frame = (hw_cap & GMAC_HW_FEAT_MGKSEL) >> 7;
> + dma_cap->pmt_remote_wake_up = (hw_cap & GMAC_HW_FEAT_RWKSEL) >> 6;
> + dma_cap->sma_mdio = (hw_cap & GMAC_HW_FEAT_SMASEL) >> 5;
> + dma_cap->vlhash = (hw_cap & GMAC_HW_FEAT_VLHASH) >> 4;
> + dma_cap->pcs = (hw_cap & GMAC_HW_FEAT_PCSSEL) >> 3;
> + dma_cap->half_duplex = (hw_cap & GMAC_HW_FEAT_HDSEL) >> 2;
> + dma_cap->mbps_1000 = (hw_cap & GMAC_HW_FEAT_GMIISEL) >> 1;
> + dma_cap->mbps_10_100 = (hw_cap & GMAC_HW_FEAT_MIISEL);
>
> /* MAC HW feature1 */
> hw_cap = readl(ioaddr + GMAC_HW_FEATURE1);
> @@ -408,7 +408,6 @@ static int dwmac4_get_hw_feature(void __iomem *ioaddr,
> dma_cap->av = (hw_cap & GMAC_HW_FEAT_AVSEL) >> 20;
> dma_cap->tsoen = (hw_cap & GMAC_HW_TSOEN) >> 18;
> dma_cap->sphen = (hw_cap & GMAC_HW_FEAT_SPHEN) >> 17;
> -
> dma_cap->addr64 = (hw_cap & GMAC_HW_ADDR64) >> 14;
> switch (dma_cap->addr64) {
> case 0:
> @@ -424,31 +423,30 @@ static int dwmac4_get_hw_feature(void __iomem *ioaddr,
> dma_cap->addr64 = 32;
> break;
> }
> -
> /* RX and TX FIFO sizes are encoded as log2(n / 128). Undo that by
> * shifting and store the sizes in bytes.
> */
> dma_cap->tx_fifo_size = 128 << ((hw_cap & GMAC_HW_TXFIFOSIZE) >> 6);
> dma_cap->rx_fifo_size = 128 << ((hw_cap & GMAC_HW_RXFIFOSIZE) >> 0);
> +
> /* MAC HW feature2 */
> hw_cap = readl(ioaddr + GMAC_HW_FEATURE2);
> + /* Number of Auxiliary Snapshot Inputs */
> + dma_cap->aux_snapshot_n = (hw_cap & GMAC_HW_FEAT_AUXSNAPNUM) >> 28;
> + /* PPS output */
> + dma_cap->pps_out_num = (hw_cap & GMAC_HW_FEAT_PPSOUTNUM) >> 24;
> /* TX and RX number of channels */
> - dma_cap->number_rx_channel =
> - ((hw_cap & GMAC_HW_FEAT_RXCHCNT) >> 12) + 1;
> dma_cap->number_tx_channel =
> ((hw_cap & GMAC_HW_FEAT_TXCHCNT) >> 18) + 1;
> + dma_cap->number_rx_channel =
> + ((hw_cap & GMAC_HW_FEAT_RXCHCNT) >> 12) + 1;
> /* TX and RX number of queues */
> - dma_cap->number_rx_queues =
> - ((hw_cap & GMAC_HW_FEAT_RXQCNT) >> 0) + 1;
> dma_cap->number_tx_queues =
> ((hw_cap & GMAC_HW_FEAT_TXQCNT) >> 6) + 1;
> - /* PPS output */
> - dma_cap->pps_out_num = (hw_cap & GMAC_HW_FEAT_PPSOUTNUM) >> 24;
> -
> + dma_cap->number_rx_queues =
> + ((hw_cap & GMAC_HW_FEAT_RXQCNT) >> 0) + 1;
> /* IEEE 1588-2002 */
> dma_cap->time_stamp = 0;
> - /* Number of Auxiliary Snapshot Inputs */
> - dma_cap->aux_snapshot_n = (hw_cap & GMAC_HW_FEAT_AUXSNAPNUM) >> 28;
>
> /* MAC HW feature3 */
> hw_cap = readl(ioaddr + GMAC_HW_FEATURE3);
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH 1/1] net: stmmac: dwmac4: extract MAC_HW_Feature regs MSB first
2023-08-10 9:59 [PATCH 1/1] net: stmmac: dwmac4: extract MAC_HW_Feature regs MSB first Furong Xu
2023-08-10 12:55 ` Alexandre TORGUE
@ 2023-08-10 13:59 ` Andrew Lunn
2023-08-11 4:14 ` Furong Xu
1 sibling, 1 reply; 4+ messages in thread
From: Andrew Lunn @ 2023-08-10 13:59 UTC (permalink / raw)
To: Furong Xu
Cc: David S. Miller, Alexandre Torgue, Jose Abreu, Eric Dumazet,
Jakub Kicinski, Paolo Abeni, Maxime Coquelin, Joao Pinto,
Simon Horman, netdev, linux-stm32, linux-arm-kernel, linux-kernel,
xfr, rock.xu
On Thu, Aug 10, 2023 at 05:59:29PM +0800, Furong Xu wrote:
> The bit extraction sequences are a little messy.
> No code changes, just reorder them.
How is this substantially better?
If bugs are found in this code, your change is going to make it harder
to back port the fixes to stable kernels. cherry-pick is unlikely to
work, it will need a human to look at it. Not hard, but still effort
for somebody.
So i think there needs to be a clear benefit to this, and you should
mention the benefit in the commit message.
Andrew
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH 1/1] net: stmmac: dwmac4: extract MAC_HW_Feature regs MSB first
2023-08-10 13:59 ` Andrew Lunn
@ 2023-08-11 4:14 ` Furong Xu
0 siblings, 0 replies; 4+ messages in thread
From: Furong Xu @ 2023-08-11 4:14 UTC (permalink / raw)
To: Andrew Lunn
Cc: David S. Miller, Alexandre Torgue, Jose Abreu, Eric Dumazet,
Jakub Kicinski, Paolo Abeni, Maxime Coquelin, Joao Pinto,
Simon Horman, netdev, linux-stm32, linux-arm-kernel, linux-kernel,
xfr, rock.xu
On Thu, 10 Aug 2023 15:59:25 +0200
Andrew Lunn <andrew@lunn.ch> wrote:
> On Thu, Aug 10, 2023 at 05:59:29PM +0800, Furong Xu wrote:
> > The bit extraction sequences are a little messy.
> > No code changes, just reorder them.
>
> How is this substantially better?
>
> If bugs are found in this code, your change is going to make it harder
> to back port the fixes to stable kernels. cherry-pick is unlikely to
> work, it will need a human to look at it. Not hard, but still effort
> for somebody.
>
> So i think there needs to be a clear benefit to this, and you should
> mention the benefit in the commit message.
>
> Andrew
Hi Andrew, Alexandre
In Synopsys databook, the description tables about all registers
start with the MSB.
So we can read one line code and then check it against one table row,
and goto next line of code against next row of table.
I agree with that this commit make cherry-pick less compatible.
If the benefit above is not worthy enough, we abandon this commit :)
Thanks.
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2023-08-11 4:14 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2023-08-10 9:59 [PATCH 1/1] net: stmmac: dwmac4: extract MAC_HW_Feature regs MSB first Furong Xu
2023-08-10 12:55 ` Alexandre TORGUE
2023-08-10 13:59 ` Andrew Lunn
2023-08-11 4:14 ` Furong Xu
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).