* [PATCH v2] ARM: dts: aspeed: Add a palmetto board (AST2400)
@ 2022-06-08 8:22 Cédric Le Goater
2022-06-08 10:17 ` Zev Weiss
0 siblings, 1 reply; 6+ messages in thread
From: Cédric Le Goater @ 2022-06-08 8:22 UTC (permalink / raw)
To: openbmc; +Cc: Joel Stanley, Cédric Le Goater
Palmettos were the first OpenPOWER systems built by Tyan with the
product name: GN70-BP010. They contained an IBM POWER8 Turismo SCM
processor and an Aspeed AST2400 SoC.
The ast2400_openbmc_defconfig defconfig is based on the evb-ast2400
defconfig, plus these extras :
CONFIG_BOOTCOMMAND="bootm 20080000"
CONFIG_DEFAULT_DEVICE_TREE="ast2400-palmetto"
CONFIG_PHY_NCSI=y
Cc: Joel Stanley <joel@jms.id.au>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
---
Changes in v2 :
- device tree cleanups to match HW
- renamed defconfig to ast2400_openbmc_defconfig
arch/arm/dts/Makefile | 1 +
arch/arm/dts/ast2400-palmetto.dts | 74 +++++++++++++++++++++++++++++++
configs/ast2400_openbmc_defconfig | 72 ++++++++++++++++++++++++++++++
3 files changed, 147 insertions(+)
create mode 100644 arch/arm/dts/ast2400-palmetto.dts
create mode 100644 configs/ast2400_openbmc_defconfig
diff --git a/arch/arm/dts/Makefile b/arch/arm/dts/Makefile
index 8f876a0aa0d7..1d523c5d5da0 100755
--- a/arch/arm/dts/Makefile
+++ b/arch/arm/dts/Makefile
@@ -676,6 +676,7 @@ dtb-$(CONFIG_ARCH_BCM6858) += \
dtb-$(CONFIG_ARCH_ASPEED) += \
ast2400-evb.dtb \
+ ast2400-palmetto.dtb \
ast2400-ahe-50dc.dtb \
ast2500-evb.dtb \
ast2600a0-evb.dtb \
diff --git a/arch/arm/dts/ast2400-palmetto.dts b/arch/arm/dts/ast2400-palmetto.dts
new file mode 100644
index 000000000000..2a49ded27b43
--- /dev/null
+++ b/arch/arm/dts/ast2400-palmetto.dts
@@ -0,0 +1,74 @@
+// SPDX-License-Identifier: GPL-2.0+
+/dts-v1/;
+
+#include "ast2400-u-boot.dtsi"
+
+/ {
+ model = "Palmetto BMC";
+ compatible = "tyan,palmetto-bmc", "aspeed,ast2400";
+
+ memory@40000000 {
+ device_type = "memory";
+ reg = <0x40000000 0x10000000>;
+ };
+
+ chosen {
+ stdout-path = &uart5;
+ bootargs = "console=ttyS4,115200 earlyprintk";
+ };
+
+ aliases {
+ spi0 = &fmc;
+ spi1 = &spi1;
+ ethernet0 = &mac0;
+ };
+};
+
+&uart5 {
+ u-boot,dm-pre-reloc;
+ status = "okay";
+};
+
+&sdrammc {
+ clock-frequency = <200000000>;
+};
+
+&wdt1 {
+ u-boot,dm-pre-reloc;
+ status = "okay";
+};
+
+&wdt2 {
+ u-boot,dm-pre-reloc;
+ status = "okay";
+};
+
+&mac0 {
+ status = "okay";
+ phy-mode = "rgmii";
+
+ pinctrl-names = "default";
+ pinctrl-0 = <&pinctrl_mac1link_default &pinctrl_mdio1_default>;
+};
+
+&fmc {
+ status = "okay";
+ flash@0 {
+ status = "okay";
+ spi-max-frequency = <50000000>;
+ spi-tx-bus-width = <2>;
+ spi-rx-bus-width = <2>;
+ };
+};
+
+&spi1 {
+ status = "okay";
+ flash@0 {
+ pinctrl-names = "default";
+ pinctrl-0 = <&pinctrl_spi1_default>;
+ status = "okay";
+ spi-max-frequency = <50000000>;
+ spi-tx-bus-width = <2>;
+ spi-rx-bus-width = <2>;
+ };
+};
diff --git a/configs/ast2400_openbmc_defconfig b/configs/ast2400_openbmc_defconfig
new file mode 100644
index 000000000000..d09358777304
--- /dev/null
+++ b/configs/ast2400_openbmc_defconfig
@@ -0,0 +1,72 @@
+CONFIG_ARM=y
+CONFIG_ARCH_ASPEED=y
+CONFIG_SYS_TEXT_BASE=0x0
+CONFIG_ASPEED_AST2400=y
+CONFIG_SYS_MALLOC_F_LEN=0x2000
+CONFIG_ENV_SIZE=0x10000
+CONFIG_ENV_OFFSET=0xF0000
+CONFIG_NR_DRAM_BANKS=1
+CONFIG_FIT=y
+CONFIG_USE_BOOTARGS=y
+CONFIG_BOOTARGS="console=ttyS4,115200n8 root=/dev/ram rw"
+CONFIG_USE_BOOTCOMMAND=y
+CONFIG_BOOTCOMMAND="bootm 20080000"
+CONFIG_PRE_CONSOLE_BUFFER=y
+CONFIG_PRE_CON_BUF_ADDR=0x1e720000
+CONFIG_SYS_CONSOLE_ENV_OVERWRITE=y
+CONFIG_HUSH_PARSER=y
+# CONFIG_AUTO_COMPLETE is not set
+CONFIG_SYS_PROMPT="ast# "
+# CONFIG_CMD_IMI is not set
+# CONFIG_CMD_XIMG is not set
+CONFIG_CMD_MEMTEST=y
+CONFIG_SYS_ALT_MEMTEST=y
+CONFIG_CMD_CLK=y
+CONFIG_CMD_GPIO=y
+CONFIG_CMD_I2C=y
+CONFIG_CMD_MMC=y
+CONFIG_CMD_SF=y
+CONFIG_CMD_DHCP=y
+CONFIG_CMD_MII=y
+CONFIG_CMD_PING=y
+CONFIG_DEFAULT_DEVICE_TREE="ast2400-palmetto"
+CONFIG_ENV_IS_IN_SPI_FLASH=y
+CONFIG_USE_ENV_SPI_BUS=y
+CONFIG_ENV_SPI_BUS=0
+CONFIG_USE_ENV_SPI_CS=y
+CONFIG_ENV_SPI_CS=0
+CONFIG_USE_ENV_SPI_MAX_HZ=y
+CONFIG_ENV_SPI_MAX_HZ=100000000
+CONFIG_NET_RANDOM_ETHADDR=y
+CONFIG_REGMAP=y
+CONFIG_CLK=y
+CONFIG_DM_GPIO=y
+CONFIG_ASPEED_GPIO=y
+CONFIG_DM_I2C=y
+CONFIG_SYS_I2C_ASPEED=y
+CONFIG_MISC=y
+CONFIG_DM_MMC=y
+# CONFIG_MMC_HW_PARTITIONING is not set
+CONFIG_MMC_SDHCI=y
+CONFIG_MMC_SDHCI_ASPEED=y
+CONFIG_DM_SPI_FLASH=y
+CONFIG_SPI_FLASH=y
+CONFIG_SPI_FLASH_MACRONIX=y
+CONFIG_SPI_FLASH_SPANSION=y
+CONFIG_SPI_FLASH_STMICRO=y
+CONFIG_SPI_FLASH_WINBOND=y
+CONFIG_PHY_REALTEK=y
+CONFIG_PHY_NCSI=y
+CONFIG_DM_ETH=y
+CONFIG_PHY_GIGE=y
+CONFIG_FTGMAC100=y
+CONFIG_PHY=y
+CONFIG_PINCTRL=y
+CONFIG_DM_SERIAL=y
+CONFIG_SYS_NS16550=y
+CONFIG_SPI=y
+CONFIG_DM_SPI=y
+CONFIG_SYSRESET=y
+CONFIG_TIMER=y
+CONFIG_WDT=y
+# CONFIG_EFI_LOADER is not set
--
2.35.3
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH v2] ARM: dts: aspeed: Add a palmetto board (AST2400)
2022-06-08 8:22 [PATCH v2] ARM: dts: aspeed: Add a palmetto board (AST2400) Cédric Le Goater
@ 2022-06-08 10:17 ` Zev Weiss
2022-06-08 13:08 ` Cédric Le Goater
0 siblings, 1 reply; 6+ messages in thread
From: Zev Weiss @ 2022-06-08 10:17 UTC (permalink / raw)
To: Cédric Le Goater; +Cc: openbmc, Joel Stanley
On Wed, Jun 08, 2022 at 01:22:21AM PDT, Cédric Le Goater wrote:
>Palmettos were the first OpenPOWER systems built by Tyan with the
>product name: GN70-BP010. They contained an IBM POWER8 Turismo SCM
>processor and an Aspeed AST2400 SoC.
>
>The ast2400_openbmc_defconfig defconfig is based on the evb-ast2400
>defconfig, plus these extras :
>
> CONFIG_BOOTCOMMAND="bootm 20080000"
> CONFIG_DEFAULT_DEVICE_TREE="ast2400-palmetto"
> CONFIG_PHY_NCSI=y
>
>Cc: Joel Stanley <joel@jms.id.au>
>Signed-off-by: Cédric Le Goater <clg@kaod.org>
>---
>
> Changes in v2 :
>
> - device tree cleanups to match HW
> - renamed defconfig to ast2400_openbmc_defconfig
>
> arch/arm/dts/Makefile | 1 +
> arch/arm/dts/ast2400-palmetto.dts | 74 +++++++++++++++++++++++++++++++
> configs/ast2400_openbmc_defconfig | 72 ++++++++++++++++++++++++++++++
> 3 files changed, 147 insertions(+)
> create mode 100644 arch/arm/dts/ast2400-palmetto.dts
> create mode 100644 configs/ast2400_openbmc_defconfig
>
>diff --git a/arch/arm/dts/Makefile b/arch/arm/dts/Makefile
>index 8f876a0aa0d7..1d523c5d5da0 100755
>--- a/arch/arm/dts/Makefile
>+++ b/arch/arm/dts/Makefile
>@@ -676,6 +676,7 @@ dtb-$(CONFIG_ARCH_BCM6858) += \
>
> dtb-$(CONFIG_ARCH_ASPEED) += \
> ast2400-evb.dtb \
>+ ast2400-palmetto.dtb \
> ast2400-ahe-50dc.dtb \
Sort? (palmetto after ahe-50dc)
> ast2500-evb.dtb \
> ast2600a0-evb.dtb \
>diff --git a/arch/arm/dts/ast2400-palmetto.dts b/arch/arm/dts/ast2400-palmetto.dts
>new file mode 100644
>index 000000000000..2a49ded27b43
>--- /dev/null
>+++ b/arch/arm/dts/ast2400-palmetto.dts
>@@ -0,0 +1,74 @@
>+// SPDX-License-Identifier: GPL-2.0+
>+/dts-v1/;
>+
>+#include "ast2400-u-boot.dtsi"
>+
>+/ {
>+ model = "Palmetto BMC";
>+ compatible = "tyan,palmetto-bmc", "aspeed,ast2400";
>+
>+ memory@40000000 {
>+ device_type = "memory";
>+ reg = <0x40000000 0x10000000>;
>+ };
>+
>+ chosen {
>+ stdout-path = &uart5;
>+ bootargs = "console=ttyS4,115200 earlyprintk";
Does bootargs serve a particular purpose here? (The other ast2x00
device-trees don't appear to include it.)
>+ };
>+
>+ aliases {
>+ spi0 = &fmc;
>+ spi1 = &spi1;
>+ ethernet0 = &mac0;
>+ };
>+};
>+
>+&uart5 {
>+ u-boot,dm-pre-reloc;
>+ status = "okay";
>+};
>+
>+&sdrammc {
>+ clock-frequency = <200000000>;
>+};
>+
>+&wdt1 {
>+ u-boot,dm-pre-reloc;
>+ status = "okay";
>+};
>+
>+&wdt2 {
>+ u-boot,dm-pre-reloc;
>+ status = "okay";
>+};
>+
>+&mac0 {
>+ status = "okay";
>+ phy-mode = "rgmii";
>+
>+ pinctrl-names = "default";
>+ pinctrl-0 = <&pinctrl_mac1link_default &pinctrl_mdio1_default>;
>+};
>+
>+&fmc {
>+ status = "okay";
>+ flash@0 {
>+ status = "okay";
>+ spi-max-frequency = <50000000>;
>+ spi-tx-bus-width = <2>;
>+ spi-rx-bus-width = <2>;
>+ };
>+};
>+
>+&spi1 {
>+ status = "okay";
>+ flash@0 {
>+ pinctrl-names = "default";
>+ pinctrl-0 = <&pinctrl_spi1_default>;
>+ status = "okay";
>+ spi-max-frequency = <50000000>;
>+ spi-tx-bus-width = <2>;
>+ spi-rx-bus-width = <2>;
>+ };
>+};
>diff --git a/configs/ast2400_openbmc_defconfig b/configs/ast2400_openbmc_defconfig
>new file mode 100644
>index 000000000000..d09358777304
>--- /dev/null
>+++ b/configs/ast2400_openbmc_defconfig
Is the implication here that the testing discussed in the v1 thread
(using the existing evb defconfig instead of an entirely new one) didn't
work out?
>@@ -0,0 +1,72 @@
>+CONFIG_ARM=y
>+CONFIG_ARCH_ASPEED=y
>+CONFIG_SYS_TEXT_BASE=0x0
>+CONFIG_ASPEED_AST2400=y
>+CONFIG_SYS_MALLOC_F_LEN=0x2000
>+CONFIG_ENV_SIZE=0x10000
>+CONFIG_ENV_OFFSET=0xF0000
>+CONFIG_NR_DRAM_BANKS=1
>+CONFIG_FIT=y
>+CONFIG_USE_BOOTARGS=y
>+CONFIG_BOOTARGS="console=ttyS4,115200n8 root=/dev/ram rw"
>+CONFIG_USE_BOOTCOMMAND=y
>+CONFIG_BOOTCOMMAND="bootm 20080000"
>+CONFIG_PRE_CONSOLE_BUFFER=y
>+CONFIG_PRE_CON_BUF_ADDR=0x1e720000
>+CONFIG_SYS_CONSOLE_ENV_OVERWRITE=y
>+CONFIG_HUSH_PARSER=y
>+# CONFIG_AUTO_COMPLETE is not set
>+CONFIG_SYS_PROMPT="ast# "
>+# CONFIG_CMD_IMI is not set
>+# CONFIG_CMD_XIMG is not set
>+CONFIG_CMD_MEMTEST=y
>+CONFIG_SYS_ALT_MEMTEST=y
>+CONFIG_CMD_CLK=y
>+CONFIG_CMD_GPIO=y
>+CONFIG_CMD_I2C=y
>+CONFIG_CMD_MMC=y
>+CONFIG_CMD_SF=y
>+CONFIG_CMD_DHCP=y
>+CONFIG_CMD_MII=y
>+CONFIG_CMD_PING=y
>+CONFIG_DEFAULT_DEVICE_TREE="ast2400-palmetto"
>+CONFIG_ENV_IS_IN_SPI_FLASH=y
>+CONFIG_USE_ENV_SPI_BUS=y
>+CONFIG_ENV_SPI_BUS=0
>+CONFIG_USE_ENV_SPI_CS=y
>+CONFIG_ENV_SPI_CS=0
>+CONFIG_USE_ENV_SPI_MAX_HZ=y
>+CONFIG_ENV_SPI_MAX_HZ=100000000
>+CONFIG_NET_RANDOM_ETHADDR=y
>+CONFIG_REGMAP=y
>+CONFIG_CLK=y
>+CONFIG_DM_GPIO=y
>+CONFIG_ASPEED_GPIO=y
>+CONFIG_DM_I2C=y
>+CONFIG_SYS_I2C_ASPEED=y
>+CONFIG_MISC=y
>+CONFIG_DM_MMC=y
>+# CONFIG_MMC_HW_PARTITIONING is not set
>+CONFIG_MMC_SDHCI=y
>+CONFIG_MMC_SDHCI_ASPEED=y
>+CONFIG_DM_SPI_FLASH=y
>+CONFIG_SPI_FLASH=y
>+CONFIG_SPI_FLASH_MACRONIX=y
>+CONFIG_SPI_FLASH_SPANSION=y
>+CONFIG_SPI_FLASH_STMICRO=y
>+CONFIG_SPI_FLASH_WINBOND=y
>+CONFIG_PHY_REALTEK=y
>+CONFIG_PHY_NCSI=y
>+CONFIG_DM_ETH=y
>+CONFIG_PHY_GIGE=y
>+CONFIG_FTGMAC100=y
>+CONFIG_PHY=y
>+CONFIG_PINCTRL=y
>+CONFIG_DM_SERIAL=y
>+CONFIG_SYS_NS16550=y
>+CONFIG_SPI=y
>+CONFIG_DM_SPI=y
>+CONFIG_SYSRESET=y
>+CONFIG_TIMER=y
>+CONFIG_WDT=y
>+# CONFIG_EFI_LOADER is not set
>--
>2.35.3
>
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v2] ARM: dts: aspeed: Add a palmetto board (AST2400)
2022-06-08 10:17 ` Zev Weiss
@ 2022-06-08 13:08 ` Cédric Le Goater
2022-06-10 6:49 ` Joel Stanley
0 siblings, 1 reply; 6+ messages in thread
From: Cédric Le Goater @ 2022-06-08 13:08 UTC (permalink / raw)
To: Zev Weiss; +Cc: openbmc, Joel Stanley
On 6/8/22 12:17, Zev Weiss wrote:
> On Wed, Jun 08, 2022 at 01:22:21AM PDT, Cédric Le Goater wrote:
>> Palmettos were the first OpenPOWER systems built by Tyan with the
>> product name: GN70-BP010. They contained an IBM POWER8 Turismo SCM
>> processor and an Aspeed AST2400 SoC.
>>
>> The ast2400_openbmc_defconfig defconfig is based on the evb-ast2400
>> defconfig, plus these extras :
>>
>> CONFIG_BOOTCOMMAND="bootm 20080000"
>> CONFIG_DEFAULT_DEVICE_TREE="ast2400-palmetto"
>> CONFIG_PHY_NCSI=y
>>
>> Cc: Joel Stanley <joel@jms.id.au>
>> Signed-off-by: Cédric Le Goater <clg@kaod.org>
>> ---
>>
>> Changes in v2 :
>>
>> - device tree cleanups to match HW
>> - renamed defconfig to ast2400_openbmc_defconfig
>>
>> arch/arm/dts/Makefile | 1 +
>> arch/arm/dts/ast2400-palmetto.dts | 74 +++++++++++++++++++++++++++++++
>> configs/ast2400_openbmc_defconfig | 72 ++++++++++++++++++++++++++++++
>> 3 files changed, 147 insertions(+)
>> create mode 100644 arch/arm/dts/ast2400-palmetto.dts
>> create mode 100644 configs/ast2400_openbmc_defconfig
>>
>> diff --git a/arch/arm/dts/Makefile b/arch/arm/dts/Makefile
>> index 8f876a0aa0d7..1d523c5d5da0 100755
>> --- a/arch/arm/dts/Makefile
>> +++ b/arch/arm/dts/Makefile
>> @@ -676,6 +676,7 @@ dtb-$(CONFIG_ARCH_BCM6858) += \
>>
>> dtb-$(CONFIG_ARCH_ASPEED) += \
>> ast2400-evb.dtb \
>> + ast2400-palmetto.dtb \
>> ast2400-ahe-50dc.dtb \
>
> Sort? (palmetto after ahe-50dc)
yes.
>
>> ast2500-evb.dtb \
>> ast2600a0-evb.dtb \
>> diff --git a/arch/arm/dts/ast2400-palmetto.dts b/arch/arm/dts/ast2400-palmetto.dts
>> new file mode 100644
>> index 000000000000..2a49ded27b43
>> --- /dev/null
>> +++ b/arch/arm/dts/ast2400-palmetto.dts
>> @@ -0,0 +1,74 @@
>> +// SPDX-License-Identifier: GPL-2.0+
>> +/dts-v1/;
>> +
>> +#include "ast2400-u-boot.dtsi"
>> +
>> +/ {
>> + model = "Palmetto BMC";
>> + compatible = "tyan,palmetto-bmc", "aspeed,ast2400";
>> +
>> + memory@40000000 {
>> + device_type = "memory";
>> + reg = <0x40000000 0x10000000>;
>> + };
>> +
>> + chosen {
>> + stdout-path = &uart5;
>> + bootargs = "console=ttyS4,115200 earlyprintk";
>
> Does bootargs serve a particular purpose here? (The other ast2x00
> device-trees don't appear to include it.)
It is useless indeed.
>> + };
>> +
>> + aliases {
>> + spi0 = &fmc;
>> + spi1 = &spi1;
>> + ethernet0 = &mac0;
>> + };
>> +};
>> +
>> +&uart5 {
>> + u-boot,dm-pre-reloc;
>> + status = "okay";
>> +};
>> +
>> +&sdrammc {
>> + clock-frequency = <200000000>;
>> +};
>> +
>> +&wdt1 {
>> + u-boot,dm-pre-reloc;
>> + status = "okay";
>> +};
>> +
>> +&wdt2 {
>> + u-boot,dm-pre-reloc;
>> + status = "okay";
>> +};
>> +
>> +&mac0 {
>> + status = "okay";
>> + phy-mode = "rgmii";
>> +
>> + pinctrl-names = "default";
>> + pinctrl-0 = <&pinctrl_mac1link_default &pinctrl_mdio1_default>;
>> +};
>> +
>> +&fmc {
>> + status = "okay";
>> + flash@0 {
>> + status = "okay";
>> + spi-max-frequency = <50000000>;
>> + spi-tx-bus-width = <2>;
>> + spi-rx-bus-width = <2>;
>> + };
>> +};
>> +
>> +&spi1 {
>> + status = "okay";
>> + flash@0 {
>> + pinctrl-names = "default";
>> + pinctrl-0 = <&pinctrl_spi1_default>;
>> + status = "okay";
>> + spi-max-frequency = <50000000>;
>> + spi-tx-bus-width = <2>;
>> + spi-rx-bus-width = <2>;
>> + };
>> +};
>> diff --git a/configs/ast2400_openbmc_defconfig b/configs/ast2400_openbmc_defconfig
>> new file mode 100644
>> index 000000000000..d09358777304
>> --- /dev/null
>> +++ b/configs/ast2400_openbmc_defconfig
>
> Is the implication here that the testing discussed in the v1 thread
> (using the existing evb defconfig instead of an entirely new one) didn't
> work out?
No because of NCSI support. net_loop() would need some tweak to
avoid configuring NCSI if the DT has no advertise support.
Here is a tentative try below, tested on the AST2500 EVB.
Thanks,
C.
From b621c15390719197357554f52f340aaa87dc80d4 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?C=C3=A9dric=20Le=20Goater?= <clg@kaod.org>
Date: Wed, 8 Jun 2022 15:06:53 +0200
Subject: [PATCH] net: phy: ncsi: Add phy_interface_is_ncsi() helper
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
and use it to avoid configuring NCSI in net_loop() when the DT has no
support for it.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
---
include/phy.h | 2 ++
drivers/net/phy/phy.c | 7 +++++++
net/net.c | 2 +-
3 files changed, 10 insertions(+), 1 deletion(-)
diff --git a/include/phy.h b/include/phy.h
index ed21775f3bbc..a759d04038a3 100644
--- a/include/phy.h
+++ b/include/phy.h
@@ -287,6 +287,8 @@ static inline bool phy_interface_is_sgmii(struct phy_device *phydev)
phydev->interface <= PHY_INTERFACE_MODE_QSGMII;
}
+bool phy_interface_is_ncsi(void);
+
/* PHY UIDs for various PHYs that are referenced in external code */
#define PHY_UID_CS4340 0x13e51002
#define PHY_UID_CS4223 0x03e57003
diff --git a/drivers/net/phy/phy.c b/drivers/net/phy/phy.c
index daea26c451fe..a0f7b7964464 100644
--- a/drivers/net/phy/phy.c
+++ b/drivers/net/phy/phy.c
@@ -988,3 +988,10 @@ int phy_get_interface_by_name(const char *str)
return -1;
}
+
+bool phy_interface_is_ncsi(void)
+{
+ struct eth_pdata *pdata = dev_get_platdata(eth_get_dev());
+
+ return pdata->phy_interface == PHY_INTERFACE_MODE_NCSI;
+}
diff --git a/net/net.c b/net/net.c
index f0f9cd4d946b..8449445bf65c 100644
--- a/net/net.c
+++ b/net/net.c
@@ -409,7 +409,7 @@ int net_loop(enum proto_t protocol)
debug_cond(DEBUG_INT_STATE, "--- net_loop Entry\n");
#ifdef CONFIG_PHY_NCSI
- if (protocol != NCSI && !ncsi_active()) {
+ if (phy_interface_is_ncsi() && protocol != NCSI && !ncsi_active()) {
printf("Configuring NCSI\n");
if (net_loop(NCSI) < 0)
return ret;
--
2.35.3
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH v2] ARM: dts: aspeed: Add a palmetto board (AST2400)
2022-06-08 13:08 ` Cédric Le Goater
@ 2022-06-10 6:49 ` Joel Stanley
2022-06-10 7:54 ` Cédric Le Goater
0 siblings, 1 reply; 6+ messages in thread
From: Joel Stanley @ 2022-06-10 6:49 UTC (permalink / raw)
To: Cédric Le Goater; +Cc: openbmc, Zev Weiss
> No because of NCSI support. net_loop() would need some tweak to
> avoid configuring NCSI if the DT has no advertise support.
>
> Here is a tentative try below, tested on the AST2500 EVB.
Thanks for the patch. I tested it in qemu for rainier (p10bmc) and
ast2600 evb + CONFIG_PHY_NCSI=y and it worked.
I then tested on hardware and it failed, as it seems we're still
trying to talk ncsi. Here's the call stack when running 'dhcp' from
the u-boot command line:
#0 ncsi_startup (phydev=0xbcf76bc8) at ../drivers/net/phy/ncsi.c:862
#1 0xbef95e28 in ftgmac100_start (dev=<optimized out>) at
../drivers/net/ftgmac100.c:378
#2 0xbefbf488 in eth_init () at ../net/eth-uclass.c:275
#3 0xbefc0814 in net_loop (protocol=DHCP) at ../net/net.c:426
ftgmac100_phy_init is calling phy_connect, and it's connecting the
ncsi phy driver:
struct phy_device *phy_connect(struct mii_dev *bus, int addr,
struct eth_device *dev,
phy_interface_t interface)
{
struct phy_device *phydev = NULL;
uint mask = (addr > 0) ? (1 << addr) : 0xffffffff;
#ifdef CONFIG_PHY_FIXED
phydev = phy_connect_fixed(bus, dev, interface);
#endif
#ifdef CONFIG_PHY_NCSI
phydev = phy_device_create(bus, 0, PHY_NCSI_ID, false, interface);
#endif
This logic will need to be fixed so we only set ncsi if the device
tree specifies it.
> Subject: [PATCH] net: phy: ncsi: Add phy_interface_is_ncsi() helper
> --- a/drivers/net/phy/phy.c
> +++ b/drivers/net/phy/phy.c
> @@ -988,3 +988,10 @@ int phy_get_interface_by_name(const char *str)
>
> return -1;
> }
> +
> +bool phy_interface_is_ncsi(void)
> +{
> + struct eth_pdata *pdata = dev_get_platdata(eth_get_dev());
> +
> + return pdata->phy_interface == PHY_INTERFACE_MODE_NCSI;
> +}
> diff --git a/net/net.c b/net/net.c
> index f0f9cd4d946b..8449445bf65c 100644
> --- a/net/net.c
> +++ b/net/net.c
> @@ -409,7 +409,7 @@ int net_loop(enum proto_t protocol)
> debug_cond(DEBUG_INT_STATE, "--- net_loop Entry\n");
>
> #ifdef CONFIG_PHY_NCSI
> - if (protocol != NCSI && !ncsi_active()) {
> + if (phy_interface_is_ncsi() && protocol != NCSI && !ncsi_active()) {
> printf("Configuring NCSI\n");
> if (net_loop(NCSI) < 0)
> return ret;
> --
> 2.35.3
>
>
>
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v2] ARM: dts: aspeed: Add a palmetto board (AST2400)
2022-06-10 6:49 ` Joel Stanley
@ 2022-06-10 7:54 ` Cédric Le Goater
2022-06-10 7:59 ` Joel Stanley
0 siblings, 1 reply; 6+ messages in thread
From: Cédric Le Goater @ 2022-06-10 7:54 UTC (permalink / raw)
To: Joel Stanley; +Cc: openbmc, Zev Weiss
On 6/10/22 08:49, Joel Stanley wrote:
>> No because of NCSI support. net_loop() would need some tweak to
>> avoid configuring NCSI if the DT has no advertise support.
>>
>> Here is a tentative try below, tested on the AST2500 EVB.
>
> Thanks for the patch. I tested it in qemu for rainier (p10bmc) and
> ast2600 evb + CONFIG_PHY_NCSI=y and it worked.
>
> I then tested on hardware and it failed, as it seems we're still
> trying to talk ncsi. Here's the call stack when running 'dhcp' from
> the u-boot command line:
>
> #0 ncsi_startup (phydev=0xbcf76bc8) at ../drivers/net/phy/ncsi.c:862
> #1 0xbef95e28 in ftgmac100_start (dev=<optimized out>) at
> ../drivers/net/ftgmac100.c:378
> #2 0xbefbf488 in eth_init () at ../net/eth-uclass.c:275
> #3 0xbefc0814 in net_loop (protocol=DHCP) at ../net/net.c:426
>
> ftgmac100_phy_init is calling phy_connect, and it's connecting the
> ncsi phy driver:
>
> struct phy_device *phy_connect(struct mii_dev *bus, int addr,
> struct eth_device *dev,
> phy_interface_t interface)
> {
> struct phy_device *phydev = NULL;
> uint mask = (addr > 0) ? (1 << addr) : 0xffffffff;
>
> #ifdef CONFIG_PHY_FIXED
> phydev = phy_connect_fixed(bus, dev, interface);
> #endif
>
> #ifdef CONFIG_PHY_NCSI
> phydev = phy_device_create(bus, 0, PHY_NCSI_ID, false, interface);
> #endif
That's ugly :/ and we have the ncsi information in the caller.
> This logic will need to be fixed so we only set ncsi if the device
> tree specifies it.
yes. Something like :
@@ -932,7 +932,13 @@ struct phy_device *phy_connect(struct mi
#endif
#ifdef CONFIG_PHY_NCSI
- phydev = phy_device_create(bus, 0, PHY_NCSI_ID, false, interface);
+ {
+ struct eth_pdata *pdata = dev_get_platdata(dev);
+
+ if (pdata->phy_interface == PHY_INTERFACE_MODE_NCSI)
+ phydev = phy_device_create(bus, 0, PHY_NCSI_ID,
+ false, interface);
+ }
#endif
if (!phydev)
?
C.
>> Subject: [PATCH] net: phy: ncsi: Add phy_interface_is_ncsi() helper
>
>> --- a/drivers/net/phy/phy.c
>> +++ b/drivers/net/phy/phy.c
>> @@ -988,3 +988,10 @@ int phy_get_interface_by_name(const char *str)
>>
>> return -1;
>> }
>> +
>> +bool phy_interface_is_ncsi(void)
>> +{
>> + struct eth_pdata *pdata = dev_get_platdata(eth_get_dev());
>
>
>> +
>> + return pdata->phy_interface == PHY_INTERFACE_MODE_NCSI;
>> +}
>> diff --git a/net/net.c b/net/net.c
>> index f0f9cd4d946b..8449445bf65c 100644
>> --- a/net/net.c
>> +++ b/net/net.c
>> @@ -409,7 +409,7 @@ int net_loop(enum proto_t protocol)
>> debug_cond(DEBUG_INT_STATE, "--- net_loop Entry\n");
>>
>> #ifdef CONFIG_PHY_NCSI
>> - if (protocol != NCSI && !ncsi_active()) {
>> + if (phy_interface_is_ncsi() && protocol != NCSI && !ncsi_active()) {
>> printf("Configuring NCSI\n");
>> if (net_loop(NCSI) < 0)
>> return ret;
>> --
>> 2.35.3
>>
>>
>>
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v2] ARM: dts: aspeed: Add a palmetto board (AST2400)
2022-06-10 7:54 ` Cédric Le Goater
@ 2022-06-10 7:59 ` Joel Stanley
0 siblings, 0 replies; 6+ messages in thread
From: Joel Stanley @ 2022-06-10 7:59 UTC (permalink / raw)
To: Cédric Le Goater; +Cc: openbmc, Zev Weiss
On Fri, 10 Jun 2022 at 07:54, Cédric Le Goater <clg@kaod.org> wrote:
>
> On 6/10/22 08:49, Joel Stanley wrote:
> >> No because of NCSI support. net_loop() would need some tweak to
> >> avoid configuring NCSI if the DT has no advertise support.
> >>
> >> Here is a tentative try below, tested on the AST2500 EVB.
> >
> > Thanks for the patch. I tested it in qemu for rainier (p10bmc) and
> > ast2600 evb + CONFIG_PHY_NCSI=y and it worked.
> >
> > I then tested on hardware and it failed, as it seems we're still
> > trying to talk ncsi. Here's the call stack when running 'dhcp' from
> > the u-boot command line:
> >
> > #0 ncsi_startup (phydev=0xbcf76bc8) at ../drivers/net/phy/ncsi.c:862
> > #1 0xbef95e28 in ftgmac100_start (dev=<optimized out>) at
> > ../drivers/net/ftgmac100.c:378
> > #2 0xbefbf488 in eth_init () at ../net/eth-uclass.c:275
> > #3 0xbefc0814 in net_loop (protocol=DHCP) at ../net/net.c:426
> >
> > ftgmac100_phy_init is calling phy_connect, and it's connecting the
> > ncsi phy driver:
> >
> > struct phy_device *phy_connect(struct mii_dev *bus, int addr,
> > struct eth_device *dev,
> > phy_interface_t interface)
> > {
> > struct phy_device *phydev = NULL;
> > uint mask = (addr > 0) ? (1 << addr) : 0xffffffff;
> >
> > #ifdef CONFIG_PHY_FIXED
> > phydev = phy_connect_fixed(bus, dev, interface);
> > #endif
> >
> > #ifdef CONFIG_PHY_NCSI
> > phydev = phy_device_create(bus, 0, PHY_NCSI_ID, false, interface);
> > #endif
>
> That's ugly :/ and we have the ncsi information in the caller.
>
> > This logic will need to be fixed so we only set ncsi if the device
> > tree specifies it.
>
> yes. Something like :
>
> @@ -932,7 +932,13 @@ struct phy_device *phy_connect(struct mi
> #endif
>
> #ifdef CONFIG_PHY_NCSI
> - phydev = phy_device_create(bus, 0, PHY_NCSI_ID, false, interface);
> + {
> + struct eth_pdata *pdata = dev_get_platdata(dev);
> +
> + if (pdata->phy_interface == PHY_INTERFACE_MODE_NCSI)
> + phydev = phy_device_create(bus, 0, PHY_NCSI_ID,
> + false, interface);
> + }
> #endif
>
> if (!phydev)
>
>
> ?
I have something a little different. I'll send it out in a moment.
>
> C.
>
>
> >> Subject: [PATCH] net: phy: ncsi: Add phy_interface_is_ncsi() helper
> >
> >> --- a/drivers/net/phy/phy.c
> >> +++ b/drivers/net/phy/phy.c
> >> @@ -988,3 +988,10 @@ int phy_get_interface_by_name(const char *str)
> >>
> >> return -1;
> >> }
> >> +
> >> +bool phy_interface_is_ncsi(void)
> >> +{
> >> + struct eth_pdata *pdata = dev_get_platdata(eth_get_dev());
> >
> >
> >> +
> >> + return pdata->phy_interface == PHY_INTERFACE_MODE_NCSI;
> >> +}
> >> diff --git a/net/net.c b/net/net.c
> >> index f0f9cd4d946b..8449445bf65c 100644
> >> --- a/net/net.c
> >> +++ b/net/net.c
> >> @@ -409,7 +409,7 @@ int net_loop(enum proto_t protocol)
> >> debug_cond(DEBUG_INT_STATE, "--- net_loop Entry\n");
> >>
> >> #ifdef CONFIG_PHY_NCSI
> >> - if (protocol != NCSI && !ncsi_active()) {
> >> + if (phy_interface_is_ncsi() && protocol != NCSI && !ncsi_active()) {
> >> printf("Configuring NCSI\n");
> >> if (net_loop(NCSI) < 0)
> >> return ret;
> >> --
> >> 2.35.3
> >>
> >>
> >>
>
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2022-06-10 8:00 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-06-08 8:22 [PATCH v2] ARM: dts: aspeed: Add a palmetto board (AST2400) Cédric Le Goater
2022-06-08 10:17 ` Zev Weiss
2022-06-08 13:08 ` Cédric Le Goater
2022-06-10 6:49 ` Joel Stanley
2022-06-10 7:54 ` Cédric Le Goater
2022-06-10 7:59 ` Joel Stanley
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.