Re: [RFC 2/6] x86: provide platform-devices for boot-framebuffers

Hi

On Wed, Jun 26, 2013 at 10:49 PM, Stephen Warren [off-list ref] wrote:

On 06/24/2013 04:27 PM, David Herrmann wrote:

quoted

The current situation regarding boot-framebuffers (VGA, VESA/VBE, EFI) on
x86 causes troubles when loading multiple fbdev drivers. The global
"struct screen_info" does not provide any state-tracking about which
drivers use the FBs. request_mem_region() theoretically works, but
unfortunately vesafb/efifb ignore it due to quirks for broken boards.

Avoid this by creating a "platform-framebuffer" device with a pointer
to the "struct screen_info" as platform-data. Drivers can now create
platform-drivers and the driver-core will refuse multiple drivers being
active simultaneously.

We keep the screen_info available for backwards-compatibility. Drivers
can be converted in follow-up patches.

Apart from "platform-framebuffer" devices, this also introduces a
compatibility option for "simple-framebuffer" drivers which recently got
introduced for OF based systems. If CONFIG_X86_SYSFB is selected, we
try to match the screen_info against a simple-framebuffer supported
format. If we succeed, we create a "simple-framebuffer" device instead
of a platform-framebuffer.
This allows to reuse the simplefb.c driver across architectures and also
to introduce a SimpleDRM driver. There is no need to have vesafb.c,
efifb.c, simplefb.c and more just to have architecture specific quirks
in their setup-routines.

Instead, we now move the architecture specific quirks into x86-setup and
provide a generic simple-framebuffer. For backwards-compatibility (if
strange formats are used), we still allow vesafb/efifb to be loaded
simultaneously and pick up all remaining devices.

quoted

diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig

quoted

+config X86_SYSFB
+     bool "Mark VGA/VBE/EFI FB as generic system framebuffer"
+     help
+       Firmwares often provide initial graphics framebuffers so the BIOS,
+       bootloader or kernel can show basic video-output during boot for
+       user-guidance and debugging. Historically, x86 used the VESA BIOS
+       Extensions and EFI-framebuffers for this, which are mostly limited
+       to x86. However, a generic system-framebuffer initialization emerged
+       recently on some non-x86 architectures.

After this patch has been in the kernel a while, that very last won't
really be true; simplefb won't have been introduced recently. Perhaps
just delete that one sentence?

It rather belongs in the commit message, right. I will rephrase that.

quoted

+       This option, if enabled, marks VGA/VBE/EFI framebuffers as generic
+       framebuffers so the new generic system-framebuffer drivers can be
+       used on x86.
+
+       This breaks any x86-only driver like efifb, vesafb, uvesafb, which
+       will not work if this is selected.

Doesn't that imply that some form of conflicts or depends ! statement
should be added here?

There is no real conflict here. You still can use vesafb/... with this
option, but they will not pick up the device. I intend to fix these up
to use "platform-framebuffer" devices instead of globally binding to
"struct screen_info". This way, framebuffers either end up as
simple-framebuffers or platform-framebuffers. This option selects
which device they end up as.

As all non-compatible framebuffers (with incompatible pixel-formats)
always end up as "platform-framebuffer", it still makes sense to use
vesafb as fallback. Hence, I'd not introduce any "conflicts"
dependency here.
Maybe I should rephrase the warning to something that makes clear that
if this option is selected, you need simplefb.c or simpledrm to make
use of these devices.

quoted

diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile

quoted

+obj-y                                        += sysfb.o

I suspect that should be obj-$(CONFIG_X86_SYSFB) += sysfb.o.

No. This patch tries to solve two things: First of all, every
system-framebuffer now gets a "platform-framebuffer" platform-device.
Iff X86_SYSFB is selected, it additionally tries to parse the
framebuffer information as "simple-framebuffer". If it succeeds, it
created a "simple-framebuffer" object, if it doesn't, a fallback
"platform-framebuffer" is provided.

This series is missing vesafb/efifb/.. patches, which should now bind
to "platform-framebuffer" devices instead of using "struct
screen_info" directly. I intend to add these in the next round.

quoted

diff --git a/arch/x86/kernel/sysfb.c b/arch/x86/kernel/sysfb.c

quoted

+#ifdef CONFIG_X86_SYSFB

Rather than ifdef'ing the body of this file, why not create a dummy
static inline version of add_sysfb() and put that into a header file
that users include. There should be a header file to prototype the
function anyway. That way, you avoid all of the ifdefs and static inline
functions in this file.

quoted

+static bool parse_mode(const struct screen_info *si,
+                    struct simplefb_platform_data *mode)

quoted

+                     strlcpy(mode->format, f->name, sizeof(mode->format));

Per my comments about the type of mode->format, I think that could just be:

mode->format = f->name;

... since formats[] (i.e. f) isn't initdata.

quoted

+#else /* CONFIG_X86_SYSFB */
+
+static bool parse_mode(const struct screen_info *si,
+                    struct simplefb_platform_data *mode)
+{
+     return false;
+}
+
+static int create_simplefb(const struct screen_info *si,
+                        const struct simplefb_platform_data *mode)
+{
+     return -EINVAL;
+}
+
+#endif /* CONFIG_X86_SYSFB */

Following on from my ifdef comment above, I believe those versions of
those functions will always cause add_sysfb() to return -ENODEV, so you
may as well provide a static inline for add_sysfb() instead.

No. add_sysfb() is supposed to always succeed. However, if
parse_mode/create_simplefb fail, it creates a "platform-framebuffer"
as fallback. I don't see any way to avoid these ifdefs. Considering
the explanation above, could you elaborate how you think this should
work?

Thanks
David