Skip to contentSkip to conversation

ARM: dts: aspeed: Update reserved memory nodes for zaius and witherspoon

PATCH Superseded Cyril Bur 5 participants 9 messages Feb 16, 2017 – Feb 21, 2017

Subsystems (2)

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore

Cyril Bur

author
#0 lore raw

PATCH 1/1 ARM: dts: aspeed: Update reserved memory nodes for zaius and witherspoon

---
 arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts | 9 +++++++--
 arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts       | 9 +++++++--
 2 files changed, 14 insertions(+), 4 deletions(-)
arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts +7 -2
index 4b2d8aa8abb6..2d76bfee1262 100644
@@ -25,9 +25,14 @@
 		#size-cells = <1>;
 		ranges;
 
-		flash_memory: region@94000000 {
+		flash_memory: region@98000000 {
 			no-map;
-			reg = <0x94000000 0x04000000>; /* 64M */
+			reg = <0x98000000 0x04000000>; /* 64M */
+		};
+
+		vga_memory: framebuffer@9c000000 {
+			no-map;
+			reg = <0x9c000000 0x04000000>; /* 64MB */
 		};
 	};
 
arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts +7 -2
index e1c9b3f4fe44..f30288f735bc 100644
@@ -25,9 +25,14 @@
 		#size-cells = <1>;
 		ranges;
 
-		flash_memory: region@94000000 {
+		flash_memory: region@98000000 {
 			no-map;
-			reg = <0x94000000 0x04000000>; /* 64M */
+			reg = <0x98000000 0x04000000>; /* 64M */
+		};
+
+		vga_memory: framebuffer@9c000000 {
+			no-map;
+			reg = <0x9c000000 0x04000000>; /* 64MB */
 		};
 	};
 
Signature · 1 line
2.11.1
#1 lore raw

Hi Cyril,

Quoted 14 lines · Cyril Bur
---
 arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts | 9 +++++++--
 arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts       | 9 +++++++--
 2 files changed, 14 insertions(+), 4 deletions(-)
arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts +0 -0
index 4b2d8aa8abb6..2d76bfee1262 100644
@@ -25,9 +25,14 @@
                #size-cells = <1>;
                ranges;
 

arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts line 35

Quoted 10 lines
arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts +7 -2
-               flash_memory: region@94000000 {
+               flash_memory: region@98000000 {
                        no-map;
-                       reg = <0x94000000 0x04000000>; /* 64M */
+                       reg = <0x98000000 0x04000000>; /* 64M */
+               };
+
+               vga_memory: framebuffer@9c000000 {
+                       no-map;
+                       reg = <0x9c000000 0x04000000>; /* 64MB */

This assumes that Witherspoon has 512MiB RAM, then why the memory part indicates it has 1GiB?

```
        memory {
                reg = <0x80000000 0x40000000>;
        };
```
Quoted 11 lines
arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts +0 -0
                };
        };
 
arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts +0 -0
index e1c9b3f4fe44..f30288f735bc 100644
@@ -25,9 +25,14 @@
                #size-cells = <1>;
                ranges;
 

arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts line 35

Quoted 10 lines
arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts +7 -2
-               flash_memory: region@94000000 {
+               flash_memory: region@98000000 {
                        no-map;
-                       reg = <0x94000000 0x04000000>; /* 64M */
+                       reg = <0x98000000 0x04000000>; /* 64M */
+               };
+
+               vga_memory: framebuffer@9c000000 {
+                       no-map;
+                       reg = <0x9c000000 0x04000000>; /* 64MB */

As I know Google decides to not support VGA on Zaius, so maybe vga_memory is not needed to Zaius at all.

Quoted 6 lines
arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts +0 -1
                };
        };
 
--

2.11.1

author
#2 lore raw
Quoted 30 lines · Mine

Hi Cyril,

Quoted 14 lines · Cyril Bur
---
 arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts | 9 +++++++--
 arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts       | 9 +++++++--
 2 files changed, 14 insertions(+), 4 deletions(-)
arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts +0 -0
index 4b2d8aa8abb6..2d76bfee1262 100644
@@ -25,9 +25,14 @@
                #size-cells = <1>;
                ranges;
 

arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts line 35

Quoted 10 lines
arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts +7 -2
-               flash_memory: region@94000000 {
+               flash_memory: region@98000000 {
                        no-map;
-                       reg = <0x94000000 0x04000000>; /* 64M */
+                       reg = <0x98000000 0x04000000>; /* 64M */
+               };
+
+               vga_memory: framebuffer@9c000000 {
+                       no-map;
+                       reg = <0x9c000000 0x04000000>; /* 64MB */

This assumes that Witherspoon has 512MiB RAM, then why the memory part indicates it has 1GiB?

Quoted 5 lines
```
        memory {
                reg = <0x80000000 0x40000000>;
        };
```

Oops, you're correct. I'll adjust the numbers.

Quoted 12 lines
Quoted 12 lines
arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts +0 -0
                };
        };
 
arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts +0 -0
index e1c9b3f4fe44..f30288f735bc 100644
@@ -25,9 +25,14 @@
                #size-cells = <1>;
                ranges;
 
Quoted 14 lines

arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts line 35

Quoted 10 lines
arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts +7 -2
-               flash_memory: region@94000000 {
+               flash_memory: region@98000000 {
                        no-map;
-                       reg = <0x94000000 0x04000000>; /* 64M */
+                       reg = <0x98000000 0x04000000>; /* 64M */
+               };
+
+               vga_memory: framebuffer@9c000000 {
+                       no-map;
+                       reg = <0x9c000000 0x04000000>; /* 64MB */

As I know Google decides to not support VGA on Zaius, so maybe vga_memory is not needed to Zaius at all.

Happy to drop that if there won't be any VGA on the platform.

Thanks,

Cyril

Quoted 6 lines
Quoted 6 lines
arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts +0 -1
                };
        };
 
--

2.11.1

author
#3 lore raw
Quoted 66 lines · Cyril Bur
Quoted 30 lines · Mine

Hi Cyril,

Quoted 14 lines · Cyril Bur
---
 arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts | 9 +++++++--
 arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts       | 9 +++++++--
 2 files changed, 14 insertions(+), 4 deletions(-)
arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts +0 -0
index 4b2d8aa8abb6..2d76bfee1262 100644
@@ -25,9 +25,14 @@
                #size-cells = <1>;
                ranges;
 

arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts line 35

Quoted 10 lines
arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts +7 -2
-               flash_memory: region@94000000 {
+               flash_memory: region@98000000 {
                        no-map;
-                       reg = <0x94000000 0x04000000>; /* 64M */
+                       reg = <0x98000000 0x04000000>; /* 64M */
+               };
+
+               vga_memory: framebuffer@9c000000 {
+                       no-map;
+                       reg = <0x9c000000 0x04000000>; /* 64MB */

This assumes that Witherspoon has 512MiB RAM, then why the memory part indicates it has 1GiB?

Quoted 5 lines
```
        memory {
                reg = <0x80000000 0x40000000>;
        };
```

Oops, you're correct. I'll adjust the numbers.

Quoted 12 lines
Quoted 12 lines
arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts +0 -0
                };
        };
 
arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts +0 -0
index e1c9b3f4fe44..f30288f735bc 100644
@@ -25,9 +25,14 @@
                #size-cells = <1>;
                ranges;
 
Quoted 14 lines

arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts line 35

Quoted 10 lines
arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts +7 -2
-               flash_memory: region@94000000 {
+               flash_memory: region@98000000 {
                        no-map;
-                       reg = <0x94000000 0x04000000>; /* 64M */
+                       reg = <0x98000000 0x04000000>; /* 64M */
+               };
+
+               vga_memory: framebuffer@9c000000 {
+                       no-map;
+                       reg = <0x9c000000 0x04000000>; /* 64MB */

As I know Google decides to not support VGA on Zaius, so maybe vga_memory is not needed to Zaius at all.

Quoted 1 line

Happy to drop that if there won't be any VGA on the platform.

Sorry for the followup, I assume it would then make sense to put the flash_memory at the top of RAM?

Quoted 11 lines

Thanks,

Cyril

Quoted 6 lines
Quoted 6 lines
arch/arm/boot/dts/aspeed-bmc-opp-zaius.dts +0 -1
                };
        };
 
--

2.11.1

#4 lore raw
In reply to Cyril Bur

arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts line 35

Quoted 10 lines · Cyril Bur
arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts +7 -2
-		flash_memory: region@94000000 {
+		flash_memory: region@98000000 {
 			no-map;
-			reg = <0x94000000 0x04000000>; /* 64M */
+			reg = <0x98000000 0x04000000>; /* 64M */
+		};
+
+		vga_memory: framebuffer@9c000000 {
+			no-map;
+			reg = <0x9c000000 0x04000000>; /* 64MB */

Do we really have to allocate 64MB to a VGA framebuffer? We can store a 4K resolution monitor with 32bit color in 32MB, so why is it required to reserve this much memory? Between this and the PNOR memory region, now 1/4th of the BMCs memory is reserved.

Signature · 1 line
Patrick Williams
author
#5 lore raw
Quoted 16 lines · Patrick Williams

arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts line 35

Quoted 10 lines · Cyril Bur
arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts +7 -2
-		flash_memory: region@94000000 {
+		flash_memory: region@98000000 {
 			no-map;
-			reg = <0x94000000 0x04000000>; /* 64M */
+			reg = <0x98000000 0x04000000>; /* 64M */
+		};
+
+		vga_memory: framebuffer@9c000000 {
+			no-map;
+			reg = <0x9c000000 0x04000000>; /* 64MB */

Do we really have to allocate 64MB to a VGA framebuffer? We can store a 4K resolution monitor with 32bit color in 32MB, so why is it required to reserve this much memory? Between this and the PNOR memory region, now 1/4th of the BMCs memory is reserved.

The reservation of 64MB is because the hardware is capable of using 64MB, therefore the host can use that region of memory. We're playing it safe here by ensuring the BMC doesn't use any of it to avoid any possibility of the host corrupting BMC memory. This is the most generic dts for the platform so we've gone with stability over performance.

Cyril

Quoted 1 line
#6 lore raw
Quoted 18 lines · Cyril Bur
Quoted 16 lines · Patrick Williams

arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts line 35

Quoted 10 lines · Cyril Bur
arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts +7 -2
-		flash_memory: region@94000000 {
+		flash_memory: region@98000000 {
 			no-map;
-			reg = <0x94000000 0x04000000>; /* 64M */
+			reg = <0x98000000 0x04000000>; /* 64M */
+		};
+
+		vga_memory: framebuffer@9c000000 {
+			no-map;
+			reg = <0x9c000000 0x04000000>; /* 64MB */

Do we really have to allocate 64MB to a VGA framebuffer? We can store a 4K resolution monitor with 32bit color in 32MB, so why is it required to reserve this much memory? Between this and the PNOR memory region, now 1/4th of the BMCs memory is reserved.

Quoted 5 lines

The reservation of 64MB is because the hardware is capable of using 64MB, therefore the host can use that region of memory. We're playing it safe here by ensuring the BMC doesn't use any of it to avoid any possibility of the host corrupting BMC memory. This is the most generic dts for the platform so we've gone with stability over performance.

Is there anything the BMC can do to restrict how big of a memory map the host side of the VGA can open up? If there isn't we're going to have to allocate this much on all systems because we can't trust the host to "behave".

Signature · 1 line
Patrick Williams
maintainer
#7 lore raw
Quoted 25 lines · Patrick Williams
Quoted 18 lines · Cyril Bur
Quoted 12 lines · Patrick Williams

arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts line 35

Quoted 10 lines · Cyril Bur
arch/arm/boot/dts/aspeed-bmc-opp-witherspoon.dts +7 -2
-         flash_memory: region@94000000 {
+         flash_memory: region@98000000 {
                  no-map;
-                 reg = <0x94000000 0x04000000>; /* 64M */
+                 reg = <0x98000000 0x04000000>; /* 64M */
+         };
+
+         vga_memory: framebuffer@9c000000 {
+                 no-map;
+                 reg = <0x9c000000 0x04000000>; /* 64MB */
Quoted 4 lines

Do we really have to allocate 64MB to a VGA framebuffer? We can store a 4K resolution monitor with 32bit color in 32MB, so why is it required to reserve this much memory? Between this and the PNOR memory region, now 1/4th of the BMCs memory is reserved.

Quoted 5 lines

The reservation of 64MB is because the hardware is capable of using 64MB, therefore the host can use that region of memory. We're playing it safe here by ensuring the BMC doesn't use any of it to avoid any possibility of the host corrupting BMC memory. This is the most generic dts for the platform so we've gone with stability over performance.

Quoted 4 lines

Is there anything the BMC can do to restrict how big of a memory map the host side of the VGA can open up? If there isn't we're going to have to allocate this much on all systems because we can't trust the host to "behave".

Ben and Jeremy have been enhancing the host driver to be smarter. Ben, what's the state of play?

Cheers,

Joel

#8 lore raw
Quoted 6 lines · Joel Stanley
Quoted 6 lines
Quoted 5 lines

The reservation of 64MB is because the hardware is capable of using 64MB, therefore the host can use that region of memory. We're playing it safe here by ensuring the BMC doesn't use any of it to avoid any possibility of the host corrupting BMC memory. This is the most generic dts for the platform so we've gone with stability over performance.

Quoted 7 lines
Quoted 4 lines

Is there anything the BMC can do to restrict how big of a memory map the host side of the VGA can open up?  If there isn't we're going to have to allocate this much on all systems because we can't trust the host to "behave".

Ben and Jeremy have been enhancing the host driver to be smarter. Ben, what's the state of play?

That's actually an interesting one ... there is a different mem region for host VGA and BMC graphics ...

The size it can use as FB (and exposes through PCI) can be selected via a strapping of the chip.

Ideally, when we have a DT aware uBoot, we could have it adjust that property based on the value of the strap. I think we can also change it though, thus we could probably make the SCU code in Linux adjust the value based on the DT and thus have platform specific DTs reducing the amount of VRAM exposed to the host.

Cheers,
Ben.