Skip to contentSkip to conversation

x86: add coreboot framebuffer support

PATCH In review Gerd Hoffmann 4 participants 11 messages Sep 4, 2014 – Sep 9, 2014 x86

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 x86: add coreboot framebuffer support

---
 arch/x86/Kconfig           |  12 +++
 arch/x86/kernel/Makefile   |   1 +
 arch/x86/kernel/coreboot.c | 232 +++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 245 insertions(+)
 create mode 100644 arch/x86/kernel/coreboot.c
arch/x86/Kconfig +12 -0
index 778178f..3aeb038 100644
@@ -2372,6 +2372,18 @@ config X86_SYSFB
 
 	  If unsure, say Y.
 
+config COREBOOT
+	bool "coreboot"
+	depends on X86_SYSFB
+	depends on FB_SIMPLE
+	help
+	  Add coreboot framebuffer support to the linux kernel.	 Say Y here
+	  if you want the linux kernel find and use the firmware framebuffer
+	  initialized by coreboot.  Useful when using the linux kernel as
+	  coreboot payload.
+
+	  If unsure, or if you don't know what coreboot is, say N.
+
 endmenu
 
 
arch/x86/kernel/Makefile +1 -0
index ada2e2d..4b30b0e 100644
@@ -103,6 +103,7 @@ obj-$(CONFIG_UPROBES)			+= uprobes.o
 obj-y					+= sysfb.o
 obj-$(CONFIG_X86_SYSFB)			+= sysfb_simplefb.o
 obj-$(CONFIG_EFI)			+= sysfb_efi.o
+obj-$(CONFIG_COREBOOT)			+= coreboot.o
 
 obj-$(CONFIG_PERF_EVENTS)		+= perf_regs.o
 obj-$(CONFIG_TRACING)			+= tracepoint.o
arch/x86/kernel/coreboot.c new file +232 -0

232 lines not on this page · raw message

Signature · 1 line
1.8.3.1
#1 lore raw
Quoted 20 lines · Gerd Hoffmann
---
 arch/x86/Kconfig           |  12 +++
 arch/x86/kernel/Makefile   |   1 +
 arch/x86/kernel/coreboot.c | 232 +++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 245 insertions(+)
 create mode 100644 arch/x86/kernel/coreboot.c
arch/x86/Kconfig +4 -0
index 778178f..3aeb038 100644
@@ -2372,6 +2372,18 @@ config X86_SYSFB
 
 	  If unsure, say Y.
 
+config COREBOOT
+	bool "coreboot"
+	depends on X86_SYSFB
+	depends on FB_SIMPLE

arch/x86/Kconfig config X86_SYSFB · line 2388

Quoted 10 lines
arch/x86/Kconfig +8 -0
+	help
+	  Add coreboot framebuffer support to the linux kernel.	 Say Y here
+	  if you want the linux kernel find and use the firmware framebuffer
+	  initialized by coreboot.  Useful when using the linux kernel as
+	  coreboot payload.
+
+	  If unsure, or if you don't know what coreboot is, say N.
+
 endmenu
 

This should NOT be named CONFIG_COREBOOT. CONFIG_FB_COREBOOT perhaps.

Quoted 40 lines
arch/x86/Kconfig +40 -0
+
+struct cb_header {
+	u32     signature;
+	u32     header_bytes;
+	u32     header_checksum;
+	u32     table_bytes;
+	u32     table_checksum;
+	u32     table_entries;
+};
+
+#define CB_TAG_MAINBOARD        0x0003
+#define CB_TAG_VERSION          0x0004
+#define CB_TAG_FORWARD          0x0011
+#define CB_TAG_FRAMEBUFFER      0x0012
+
+struct cb_mainboard {
+	u8      vendor_idx;
+	u8      part_idx;
+	char    strings[0];
+};
+
+struct cb_framebuffer {
+	u64     physical_address;
+	u32     x_resolution;
+	u32     y_resolution;
+	u32     bytes_per_line;
+	u8      bits_per_pixel;
+	u8      red_mask_pos;
+	u8      red_mask_size;
+	u8      green_mask_pos;
+	u8      green_mask_size;
+	u8      blue_mask_pos;
+	u8      blue_mask_size;
+	u8      reserved_mask_pos;
+	u8      reserved_mask_size;
+};
+
+struct cb_entry {
+	u32     tag;
+	u32     size;

arch/x86/Kconfig config X86_SYSFB

Quoted 10 lines
arch/x86/Kconfig +10 -0
+	union {
+		char    string[0];
+		u64     forward;
+		struct cb_mainboard   mb;
+		struct cb_framebuffer fb;
+	} u;
+};
+
+#define CB_SIGNATURE 0x4f49424C  /* "LBIO" */
+

This stuff belongs in a header file.

I'm not a fan of Coreboot having invented its own nonstandard hacks, but I guess it is pretty much unavoidable.

-hpa

#2 lore raw
Quoted 2 lines · H. Peter Anvin

I'm not a fan of Coreboot having invented its own nonstandard hacks, but I guess it is pretty much unavoidable.

I suspect this should not be architecture-dependent, since coreboot tables work across 3 coreboot architectures at present with more on the way.

Sometimes, something that spans more than an x86 can look like a non-standard hack. At other times, it might look portable. Depends on ones perspective I guess.

Gerd, I'll get back to you offline to show you what we did in Akaros.

ron

#3 lore raw
In reply to H. Peter Anvin
Quoted 3 lines

I'm not a fan of Coreboot having invented its own nonstandard hacks, but I guess it is pretty much unavoidable.

It's completely avoidable. The stub can copy this information to standard framebuffer info structure. The only missing thing is to apply patch by cjwatson or mjg59 (I'm not sure now who wrote it) for having an ID for linear framebuffer which implies no specific hardware (it can be any) or firmware (coreboot doesn't provide nay additional info or callback).

Please don't apply this patch it will break SeaBIOS booting with VGABIOS.

#4 lore raw

Vladimir can you point me to that patch? This sounds interesting.

ron

#6 lore raw
Quoted 6 lines · Vladimir 'φ-coder/phcoder' Serbinenko
Quoted 3 lines · ron minnich

Vladimir can you point me to that patch? This sounds interesting.

https://lkml.org/lkml/2010/8/25/190

I believe *most* of this patch has already gotten merged as we now use simplefb on x86 as well. So all we need is probably the ID.

-hpa

#7 lore raw
Quoted 11 lines · H. Peter Anvin
Quoted 6 lines · Vladimir 'φ-coder/phcoder' Serbinenko
Quoted 3 lines · ron minnich

Vladimir can you point me to that patch? This sounds interesting.

https://lkml.org/lkml/2010/8/25/190

I believe *most* of this patch has already gotten merged as we now use simplefb on x86 as well. So all we need is probably the ID.

Yes, efifb has the same semantics. We just need some ID with clear documentation saying sth like "implies framebuffer without anything else" so that noone will get an idea to plug e.g. vga hooks into it.

Quoted 3 lines

-hpa

#8 lore raw
Quoted 10 lines · Vladimir 'φ-coder/phcoder' Serbinenko
Quoted 9 lines · H. Peter Anvin
Quoted 6 lines · Vladimir 'φ-coder/phcoder' Serbinenko
Quoted 3 lines · ron minnich

Vladimir can you point me to that patch? This sounds interesting.

https://lkml.org/lkml/2010/8/25/190

Quoted 9 lines
Quoted 4 lines

I believe *most* of this patch has already gotten merged as we now use simplefb on x86 as well. So all we need is probably the ID.

Yes, efifb has the same semantics. We just need some ID with clear documentation saying sth like "implies framebuffer without anything else" so that noone will get an idea to plug e.g. vga hooks into it.

Want to make a patch?

-hpa

#9 lore raw

Hi,

Quoted 7 lines
Quoted 5 lines

Yes, efifb has the same semantics. We just need some ID with clear documentation saying sth like "implies framebuffer without anything else" so that noone will get an idea to plug e.g. vga hooks into it.

Want to make a patch?

Attached. Can someone test with the patched coreboot trampoline please?

cheers,
Gerd

Quoted 1 line

From cadf8315cb5d76a89a1b84e7ea22565f6b1945dc Mon Sep 17 00:00:00 2001

From: Gerd Hoffmann <[email protected]>
Date: Tue, 9 Sep 2014 08:10:09 +0200
Subject: [PATCH] Add VIDEO_TYPE_LINEAR

Add VIDEO_TYPE_LINEAR for a simple linear framebuffer, initialized by the firmware (for example coreboot).

Needs CONFIG_X86_SYSFB=y and CONFIG_FB_SIMPLE=y to work.

---
 arch/x86/kernel/sysfb_simplefb.c | 4 +++-
 include/uapi/linux/screen_info.h | 2 ++
 2 files changed, 5 insertions(+), 1 deletion(-)
arch/x86/kernel/sysfb_simplefb.c +3 -1
index 86179d4..0742c82 100644
@@ -36,7 +36,9 @@ __init bool parse_mode(const struct screen_info *si,
 	unsigned int i;
 
 	type = si->orig_video_isVGA;
-	if (type != VIDEO_TYPE_VLFB && type != VIDEO_TYPE_EFI)
+	if (type != VIDEO_TYPE_VLFB &&
+	    type != VIDEO_TYPE_EFI &&
+	    type != VIDEO_TYPE_LINEAR)
 		return false;
 
 	for (i = 0; i < ARRAY_SIZE(formats); ++i) {
include/uapi/linux/screen_info.h +2 -0
index 7530e74..d27126d 100644
@@ -66,6 +66,8 @@ struct screen_info {
 
 #define VIDEO_TYPE_EFI		0x70	/* EFI graphic mode		*/
 
+#define VIDEO_TYPE_LINEAR	0x80	/* simple linear framebuffer	*/
+
 #define VIDEO_FLAGS_NOCURSOR	(1 << 0) /* The video mode has no cursor set */
 
 #define VIDEO_CAPABILITY_SKIP_QUIRKS	(1 << 0)
Signature · 1 line
1.8.3.1
#10 lore raw
In reply to Vladimir 'φ-coder/phcoder' Serbinenko
Quoted 10 lines · Vladimir 'φ-coder/phcoder' Serbinenko
Quoted 3 lines

I'm not a fan of Coreboot having invented its own nonstandard hacks, but I guess it is pretty much unavoidable.

It's completely avoidable. The stub can copy this information to standard framebuffer info structure. The only missing thing is to apply patch by cjwatson or mjg59 (I'm not sure now who wrote it) for having an ID for linear framebuffer which implies no specific hardware (it can be any) or firmware (coreboot doesn't provide nay additional info or callback).

Quoted 3 lines

Please don't apply this patch it will break SeaBIOS booting with VGABIOS.

Thanks for the warning. Yes, this is the Right Thing.

-hpa