[PATCH u-boot v2019.04-aspeed-openbmc v4 0/2] Add support for IBM Genesis3

[PATCH u-boot v2019.04-aspeed-openbmc v4 0/2] Add support for IBM Genesis3

[Patrick Rudolph]

v4:
  - Rebase on upstream
  - Drop defconfig and add it on openbmc instead
  - Rename Kconfig for debug uart

v3:
  - Drop FIRMWARE_2ND_BOOT patch
  - Include reference board DTS instead of copying it
  - Rename DTS to ast2500-<boardname>
  - Describe defconfig changes in commit message

v2:
  - Fix whitespace and compatible in DTS
  - Remove board stub
  - Improve commit message
  - Rewrite FIRMWARE_2ND_BOOT support
  - Update defconfig

Patrick Rudolph (2):
  arm/dts: Add IBM Genesis3 board
  arm/mach-aspeed: Add support for CONFIG_ASPEED_DEBUG_UART_TO_UART1

 arch/arm/dts/Makefile                   |  1 +
 arch/arm/dts/ast2500-genesis3.dts       | 28 +++++++++++++++++++++++++
 arch/arm/mach-aspeed/Kconfig            |  5 +++++
 arch/arm/mach-aspeed/ast2500/platform.S |  2 +-
 4 files changed, 35 insertions(+), 1 deletion(-)
 create mode 100644 arch/arm/dts/ast2500-genesis3.dts

-- 
2.35.3

[PATCH u-boot v2019.04-aspeed-openbmc v4 1/2] arm/dts: Add IBM Genesis3 board

[Patrick Rudolph]

Add devicetree source file. It uses the evb-ast2500 board files.

Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
---
 arch/arm/dts/Makefile             |  1 +
 arch/arm/dts/ast2500-genesis3.dts | 28 ++++++++++++++++++++++++++++
 2 files changed, 29 insertions(+)
 create mode 100644 arch/arm/dts/ast2500-genesis3.dts

diff --git a/arch/arm/dts/Makefile b/arch/arm/dts/Makefile
index 3515100c65..ed9714d832 100755
--- a/arch/arm/dts/Makefile
+++ b/arch/arm/dts/Makefile
@@ -678,6 +678,7 @@ dtb-$(CONFIG_ARCH_ASPEED) += \
 	ast2400-evb.dtb \
 	ast2400-ahe-50dc.dtb \
 	ast2500-evb.dtb \
+	ast2500-genesis3.dtb \
 	ast2600a0-evb.dtb \
 	ast2600a1-evb.dtb \
 	ast2600-bletchley.dtb \
diff --git a/arch/arm/dts/ast2500-genesis3.dts b/arch/arm/dts/ast2500-genesis3.dts
new file mode 100644
index 0000000000..544758c5b8
--- /dev/null
+++ b/arch/arm/dts/ast2500-genesis3.dts
@@ -0,0 +1,28 @@
+// SPDX-License-Identifier: GPL-2.0+
+/*
+ * IBM Genesis3
+ *
+ * Copyright (C) 2022 9elements GmbH
+ */
+
+#include "ast2500-evb.dts"
+
+/ {
+	model = "IBM Genesis3";
+	compatible = "ibm,genesis3-bmc", "aspeed,ast2500";
+};
+
+&spi1 {
+	status = "disabled";
+};
+
+&fmc {
+	flash@0 {
+		compatible = "spansion,s25fl256l", "spi-flash";
+	};
+
+	flash@1 {
+		compatible = "spansion,s25fl256l", "spi-flash";
+	};
+};
+
-- 
2.35.3

[PATCH u-boot v2019.04-aspeed-openbmc v4 2/2] arm/mach-aspeed: Add support for CONFIG_ASPEED_DEBUG_UART_TO_UART1

[Patrick Rudolph]

Introduce CONFIG_ASPEED_DEBUG_UART_TO_UART1 and reuse existing
platform code to route the debug uart to RDX1/TDX1.
This is required on IBM/Genesis3 as it uses RDX1/TDX1 as debug uart.

Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Reviewed-by: Joel Stanley <joel@jms.id.au>
---
 arch/arm/mach-aspeed/Kconfig            | 5 +++++
 arch/arm/mach-aspeed/ast2500/platform.S | 2 +-
 2 files changed, 6 insertions(+), 1 deletion(-)

diff --git a/arch/arm/mach-aspeed/Kconfig b/arch/arm/mach-aspeed/Kconfig
index edb5520aec..a38f070381 100644
--- a/arch/arm/mach-aspeed/Kconfig
+++ b/arch/arm/mach-aspeed/Kconfig
@@ -82,6 +82,11 @@ config ASPEED_ENABLE_DEBUG_UART
 	  systems, but may be useful to enable for debugging during
 	  development.
 
+config ASPEED_DEBUG_UART_TO_UART1
+	bool "Route debug UART to UART1"
+	depends on ASPEED_AST2500
+	help
+	  Route debug UART to TXD1/RXD1 pins.
 endif
 
 config ASPEED_PALLADIUM
diff --git a/arch/arm/mach-aspeed/ast2500/platform.S b/arch/arm/mach-aspeed/ast2500/platform.S
index aef55c4a0a..a97ebebcca 100644
--- a/arch/arm/mach-aspeed/ast2500/platform.S
+++ b/arch/arm/mach-aspeed/ast2500/platform.S
@@ -795,7 +795,7 @@ wait_ddr_reset:
     /* end delay 10ms */
 
 /* Debug - UART console message */
-#ifdef CONFIG_DRAM_UART_TO_UART1
+#ifdef CONFIG_ASPEED_DEBUG_UART_TO_UART1
     ldr   r0, =0x1e78909c                        @ route UART5 to UART Port1, 2016.08.29
     ldr   r1, =0x10000004
     str   r1, [r0]
-- 
2.35.3

Re: [PATCH u-boot v2019.04-aspeed-openbmc v4 1/2] arm/dts: Add IBM Genesis3 board

[Zev Weiss]

On Mon, May 23, 2022 at 06:40:03AM PDT, Patrick Rudolph wrote:
>Add devicetree source file. It uses the evb-ast2500 board files.
>
>Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
>---
> arch/arm/dts/Makefile             |  1 +
> arch/arm/dts/ast2500-genesis3.dts | 28 ++++++++++++++++++++++++++++
> 2 files changed, 29 insertions(+)
> create mode 100644 arch/arm/dts/ast2500-genesis3.dts
>

Reviewed-by: Zev Weiss <zweiss@equinix.com>

Re: [PATCH u-boot v2019.04-aspeed-openbmc v4 2/2] arm/mach-aspeed: Add support for CONFIG_ASPEED_DEBUG_UART_TO_UART1

[Zev Weiss]

On Mon, May 23, 2022 at 06:40:04AM PDT, Patrick Rudolph wrote:
>Introduce CONFIG_ASPEED_DEBUG_UART_TO_UART1 and reuse existing
>platform code to route the debug uart to RDX1/TDX1.
>This is required on IBM/Genesis3 as it uses RDX1/TDX1 as debug uart.
>
>Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
>Reviewed-by: Joel Stanley <joel@jms.id.au>
>---
> arch/arm/mach-aspeed/Kconfig            | 5 +++++
> arch/arm/mach-aspeed/ast2500/platform.S | 2 +-
> 2 files changed, 6 insertions(+), 1 deletion(-)
>
>diff --git a/arch/arm/mach-aspeed/Kconfig b/arch/arm/mach-aspeed/Kconfig
>index edb5520aec..a38f070381 100644
>--- a/arch/arm/mach-aspeed/Kconfig
>+++ b/arch/arm/mach-aspeed/Kconfig
>@@ -82,6 +82,11 @@ config ASPEED_ENABLE_DEBUG_UART
> 	  systems, but may be useful to enable for debugging during
> 	  development.
>
>+config ASPEED_DEBUG_UART_TO_UART1
>+	bool "Route debug UART to UART1"
>+	depends on ASPEED_AST2500
>+	help
>+	  Route debug UART to TXD1/RXD1 pins.

Any reason not to put this in 'if ASPEED_ENABLE_DEBUG_UART' as suggested
in the previous review?  And since that one already has the
ASPEED_AST2500 dependency, I think it'd obviate the need to have that
specified on ASPEED_DEBUG_UART_TO_UART1.

While we're at it, slightly more detail in the help text would good I
think, perhaps just "... instead of the default TXD5/RXD5."

Though actually, looking at the datasheet I'm now not certain if this
does exactly what I had been thinking previously -- if I'm understanding
it right, it's not so much switching the debug-UART functionality from
UART5 to UART1, it's re-routing UART5 itself to the I/Os typically used
for UART1?  Which seems somewhat different, and I guess would ultimately
be independent of the debug-UART itself being enabled or disabled, in
which case maybe what I said earlier wasn't entirely appropriate...maybe
someone with more expertise on the ast2500 UARTs (e.g. Andrew?) can
weigh in?

> endif
>
> config ASPEED_PALLADIUM
>diff --git a/arch/arm/mach-aspeed/ast2500/platform.S b/arch/arm/mach-aspeed/ast2500/platform.S
>index aef55c4a0a..a97ebebcca 100644
>--- a/arch/arm/mach-aspeed/ast2500/platform.S
>+++ b/arch/arm/mach-aspeed/ast2500/platform.S
>@@ -795,7 +795,7 @@ wait_ddr_reset:
>     /* end delay 10ms */
>
> /* Debug - UART console message */
>-#ifdef CONFIG_DRAM_UART_TO_UART1
>+#ifdef CONFIG_ASPEED_DEBUG_UART_TO_UART1
>     ldr   r0, =0x1e78909c                        @ route UART5 to UART Port1, 2016.08.29
>     ldr   r1, =0x10000004
>     str   r1, [r0]
>-- 
>2.35.3
>

Re: [PATCH u-boot v2019.04-aspeed-openbmc v4 2/2] arm/mach-aspeed: Add support for CONFIG_ASPEED_DEBUG_UART_TO_UART1

[Patrick Rudolph]

[-- Attachment #1: Type: text/plain, Size: 2926 bytes --]

On Tue, May 24, 2022 at 1:07 AM Zev Weiss <zweiss@equinix.com> wrote:

> On Mon, May 23, 2022 at 06:40:04AM PDT, Patrick Rudolph wrote:
> >Introduce CONFIG_ASPEED_DEBUG_UART_TO_UART1 and reuse existing
> >platform code to route the debug uart to RDX1/TDX1.
> >This is required on IBM/Genesis3 as it uses RDX1/TDX1 as debug uart.
> >
> >Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
> >Reviewed-by: Joel Stanley <joel@jms.id.au>
> >---
> > arch/arm/mach-aspeed/Kconfig            | 5 +++++
> > arch/arm/mach-aspeed/ast2500/platform.S | 2 +-
> > 2 files changed, 6 insertions(+), 1 deletion(-)
> >
> >diff --git a/arch/arm/mach-aspeed/Kconfig b/arch/arm/mach-aspeed/Kconfig
> >index edb5520aec..a38f070381 100644
> >--- a/arch/arm/mach-aspeed/Kconfig
> >+++ b/arch/arm/mach-aspeed/Kconfig
> >@@ -82,6 +82,11 @@ config ASPEED_ENABLE_DEBUG_UART
> >         systems, but may be useful to enable for debugging during
> >         development.
> >
> >+config ASPEED_DEBUG_UART_TO_UART1
> >+      bool "Route debug UART to UART1"
> >+      depends on ASPEED_AST2500
> >+      help
> >+        Route debug UART to TXD1/RXD1 pins.
>
> Any reason not to put this in 'if ASPEED_ENABLE_DEBUG_UART' as suggested
> in the previous review?  And since that one already has the
> ASPEED_AST2500 dependency, I think it'd obviate the need to have that
> specified on ASPEED_DEBUG_UART_TO_UART1.
>
> While we're at it, slightly more detail in the help text would good I
> think, perhaps just "... instead of the default TXD5/RXD5."
>
> Though actually, looking at the datasheet I'm now not certain if this
> does exactly what I had been thinking previously -- if I'm understanding
> it right, it's not so much switching the debug-UART functionality from
> UART5 to UART1, it's re-routing UART5 itself to the I/Os typically used
> for UART1?  Which seems somewhat different, and I guess would ultimately
> be independent of the debug-UART itself being enabled or disabled, in
> which case maybe what I said earlier wasn't entirely appropriate...maybe
> someone with more expertise on the ast2500 UARTs (e.g. Andrew?) can
> weigh in?
>
> As I understand it only re-routes the UART5 to UART1 I/Os. However it
doesn't make any
sense to re-route the UART5 if it's disabled.
I'll push a new patch.

> endif
> >
> > config ASPEED_PALLADIUM
> >diff --git a/arch/arm/mach-aspeed/ast2500/platform.S
> b/arch/arm/mach-aspeed/ast2500/platform.S
> >index aef55c4a0a..a97ebebcca 100644
> >--- a/arch/arm/mach-aspeed/ast2500/platform.S
> >+++ b/arch/arm/mach-aspeed/ast2500/platform.S
> >@@ -795,7 +795,7 @@ wait_ddr_reset:
> >     /* end delay 10ms */
> >
> > /* Debug - UART console message */
> >-#ifdef CONFIG_DRAM_UART_TO_UART1
> >+#ifdef CONFIG_ASPEED_DEBUG_UART_TO_UART1
> >     ldr   r0, =0x1e78909c                        @ route UART5 to UART
> Port1, 2016.08.29
> >     ldr   r1, =0x10000004
> >     str   r1, [r0]
> >--
> >2.35.3
> >

[-- Attachment #2: Type: text/html, Size: 3957 bytes --]

Re: [PATCH u-boot v2019.04-aspeed-openbmc v4 2/2] arm/mach-aspeed: Add support for CONFIG_ASPEED_DEBUG_UART_TO_UART1

[Zev Weiss]

On Tue, May 24, 2022 at 02:30:02AM PDT, Patrick Rudolph wrote:
> On Tue, May 24, 2022 at 1:07 AM Zev Weiss <zweiss@equinix.com> wrote:
> 
> > On Mon, May 23, 2022 at 06:40:04AM PDT, Patrick Rudolph wrote:
> > >Introduce CONFIG_ASPEED_DEBUG_UART_TO_UART1 and reuse existing
> > >platform code to route the debug uart to RDX1/TDX1.
> > >This is required on IBM/Genesis3 as it uses RDX1/TDX1 as debug uart.
> > >
> > >Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
> > >Reviewed-by: Joel Stanley <joel@jms.id.au>
> > >---
> > > arch/arm/mach-aspeed/Kconfig            | 5 +++++
> > > arch/arm/mach-aspeed/ast2500/platform.S | 2 +-
> > > 2 files changed, 6 insertions(+), 1 deletion(-)
> > >
> > >diff --git a/arch/arm/mach-aspeed/Kconfig b/arch/arm/mach-aspeed/Kconfig
> > >index edb5520aec..a38f070381 100644
> > >--- a/arch/arm/mach-aspeed/Kconfig
> > >+++ b/arch/arm/mach-aspeed/Kconfig
> > >@@ -82,6 +82,11 @@ config ASPEED_ENABLE_DEBUG_UART
> > >         systems, but may be useful to enable for debugging during
> > >         development.
> > >
> > >+config ASPEED_DEBUG_UART_TO_UART1
> > >+      bool "Route debug UART to UART1"
> > >+      depends on ASPEED_AST2500
> > >+      help
> > >+        Route debug UART to TXD1/RXD1 pins.
> >
> > Any reason not to put this in 'if ASPEED_ENABLE_DEBUG_UART' as suggested
> > in the previous review?  And since that one already has the
> > ASPEED_AST2500 dependency, I think it'd obviate the need to have that
> > specified on ASPEED_DEBUG_UART_TO_UART1.
> >
> > While we're at it, slightly more detail in the help text would good I
> > think, perhaps just "... instead of the default TXD5/RXD5."
> >
> > Though actually, looking at the datasheet I'm now not certain if this
> > does exactly what I had been thinking previously -- if I'm understanding
> > it right, it's not so much switching the debug-UART functionality from
> > UART5 to UART1, it's re-routing UART5 itself to the I/Os typically used
> > for UART1?  Which seems somewhat different, and I guess would ultimately
> > be independent of the debug-UART itself being enabled or disabled, in
> > which case maybe what I said earlier wasn't entirely appropriate...maybe
> > someone with more expertise on the ast2500 UARTs (e.g. Andrew?) can
> > weigh in?
> >
> > As I understand it only re-routes the UART5 to UART1 I/Os. However it
> doesn't make any
> sense to re-route the UART5 if it's disabled.
> I'll push a new patch.
> 

Bear in mind that there's a difference between UART5 and the debug-UART 
feature that can be enabled -- UART5 is a UART pretty much like all the 
others, but it happens to be the one on which the debug feature operates 
by default (listening for a secret password and providing access to 
various special operations via a simple command-line interface).  Even 
with CONFIG_ASPEED_ENABLE_DEBUG_UART=n, UART5 is still available for use 
as a normal UART, it just doesn't have that magic debug capability 
turned on.


Zev

Re: [PATCH u-boot v2019.04-aspeed-openbmc v4 2/2] arm/mach-aspeed: Add support for CONFIG_ASPEED_DEBUG_UART_TO_UART1

[Patrick Rudolph]

[-- Attachment #1: Type: text/plain, Size: 3348 bytes --]

On Tue, May 24, 2022 at 11:35 AM Zev Weiss <zev@bewilderbeest.net> wrote:

> On Tue, May 24, 2022 at 02:30:02AM PDT, Patrick Rudolph wrote:
> > On Tue, May 24, 2022 at 1:07 AM Zev Weiss <zweiss@equinix.com> wrote:
> >
> > > On Mon, May 23, 2022 at 06:40:04AM PDT, Patrick Rudolph wrote:
> > > >Introduce CONFIG_ASPEED_DEBUG_UART_TO_UART1 and reuse existing
> > > >platform code to route the debug uart to RDX1/TDX1.
> > > >This is required on IBM/Genesis3 as it uses RDX1/TDX1 as debug uart.
> > > >
> > > >Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
> > > >Reviewed-by: Joel Stanley <joel@jms.id.au>
> > > >---
> > > > arch/arm/mach-aspeed/Kconfig            | 5 +++++
> > > > arch/arm/mach-aspeed/ast2500/platform.S | 2 +-
> > > > 2 files changed, 6 insertions(+), 1 deletion(-)
> > > >
> > > >diff --git a/arch/arm/mach-aspeed/Kconfig
> b/arch/arm/mach-aspeed/Kconfig
> > > >index edb5520aec..a38f070381 100644
> > > >--- a/arch/arm/mach-aspeed/Kconfig
> > > >+++ b/arch/arm/mach-aspeed/Kconfig
> > > >@@ -82,6 +82,11 @@ config ASPEED_ENABLE_DEBUG_UART
> > > >         systems, but may be useful to enable for debugging during
> > > >         development.
> > > >
> > > >+config ASPEED_DEBUG_UART_TO_UART1
> > > >+      bool "Route debug UART to UART1"
> > > >+      depends on ASPEED_AST2500
> > > >+      help
> > > >+        Route debug UART to TXD1/RXD1 pins.
> > >
> > > Any reason not to put this in 'if ASPEED_ENABLE_DEBUG_UART' as
> suggested
> > > in the previous review?  And since that one already has the
> > > ASPEED_AST2500 dependency, I think it'd obviate the need to have that
> > > specified on ASPEED_DEBUG_UART_TO_UART1.
> > >
> > > While we're at it, slightly more detail in the help text would good I
> > > think, perhaps just "... instead of the default TXD5/RXD5."
> > >
> > > Though actually, looking at the datasheet I'm now not certain if this
> > > does exactly what I had been thinking previously -- if I'm
> understanding
> > > it right, it's not so much switching the debug-UART functionality from
> > > UART5 to UART1, it's re-routing UART5 itself to the I/Os typically used
> > > for UART1?  Which seems somewhat different, and I guess would
> ultimately
> > > be independent of the debug-UART itself being enabled or disabled, in
> > > which case maybe what I said earlier wasn't entirely
> appropriate...maybe
> > > someone with more expertise on the ast2500 UARTs (e.g. Andrew?) can
> > > weigh in?
> > >
> > > As I understand it only re-routes the UART5 to UART1 I/Os. However it
> > doesn't make any
> > sense to re-route the UART5 if it's disabled.
> > I'll push a new patch.
> >
>
> Bear in mind that there's a difference between UART5 and the debug-UART
> feature that can be enabled -- UART5 is a UART pretty much like all the
> others, but it happens to be the one on which the debug feature operates
> by default (listening for a secret password and providing access to
> various special operations via a simple command-line interface).  Even
> with CONFIG_ASPEED_ENABLE_DEBUG_UART=n, UART5 is still available for use
> as a normal UART, it just doesn't have that magic debug capability
> turned on.
>
> Oh, I didn't know that. With that knowledge there should be no need to
depend on the
CONFIG_ASPEED_ENABLE_DEBUG_UART when re-routing uart5.
I'll try another aproach.

>
> Zev
>
>

[-- Attachment #2: Type: text/html, Size: 4682 bytes --]