[PATCH 0/2] x86/video: improve early video detection

[PATCH 0/2] x86/video: improve early video detection

[Roger Pau Monne]

Hello,

The current video logic reports wrong values when booted in PVH mode, as
that boot path explicitly skips real-mode logic, and thus avoids any VGA
video detection.

First patch is cleanup of the offset declarations in boot_video_info.
Second patch attempts to fix Xen in PVH mode reporting VGA support.

Roger Pau Monne (2):
  x86/video: add boot_video_info offset generation to asm-offsets
  x86/video: do not assume a video mode to be unconditionally present

 xen/arch/x86/boot/video.S         | 88 +++++++++++--------------------
 xen/arch/x86/x86_64/asm-offsets.c | 26 +++++++++
 2 files changed, 58 insertions(+), 56 deletions(-)

-- 
2.44.0

[PATCH 1/2] x86/video: add boot_video_info offset generation to asm-offsets

[Roger Pau Monne]

Currently the offsets into the boot_video_info struct are manually encoded in
video.S, which is fragile.  Generate them in asm-offsets.c and switch the
current code to use those instead.

No functional change intended.

Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
---
 xen/arch/x86/boot/video.S         | 83 ++++++++++++-------------------
 xen/arch/x86/x86_64/asm-offsets.c | 26 ++++++++++
 2 files changed, 58 insertions(+), 51 deletions(-)

diff --git a/xen/arch/x86/boot/video.S b/xen/arch/x86/boot/video.S
index 0ae04f270f8c..a4b25a3b34d1 100644
--- a/xen/arch/x86/boot/video.S
+++ b/xen/arch/x86/boot/video.S
@@ -26,32 +26,13 @@
 /* Force 400 scan lines for standard modes (hack to fix bad BIOS behaviour */
 #undef CONFIG_VIDEO_400_HACK
 
-/* Positions of various video parameters passed to the kernel */
-/* (see also include/linux/tty.h) */
-#define PARAM_CURSOR_POS        0x00
-#define PARAM_VIDEO_MODE        0x02
-#define PARAM_VIDEO_COLS        0x03
-#define PARAM_VIDEO_LINES       0x04
-#define PARAM_HAVE_VGA          0x05
-#define PARAM_FONT_POINTS       0x06
-#define PARAM_CAPABILITIES      0x08
-#define PARAM_LFB_LINELENGTH    0x0c
-#define PARAM_LFB_WIDTH         0x0e
-#define PARAM_LFB_HEIGHT        0x10
-#define PARAM_LFB_DEPTH         0x12
-#define PARAM_LFB_BASE          0x14
-#define PARAM_LFB_SIZE          0x18
-#define PARAM_LFB_COLORS        0x1c
-#define PARAM_VESAPM_SEG        0x24
-#define PARAM_VESAPM_OFF        0x26
-#define PARAM_VESA_ATTRIB       0x28
 #define _param(param) bootsym(boot_vid_info)+(param)
 
 video:  xorw    %ax, %ax
         movw    %ax, %gs        # GS is zero
         cld
         call    basic_detect    # Basic adapter type testing (EGA/VGA/MDA/CGA)
-        cmpb    $0,_param(PARAM_HAVE_VGA)
+        cmpb    $0,_param(BVI_have_vga)
         je      1f                # Bail if there's no VGA
         movw    bootsym(boot_vid_mode), %ax     # User selected video mode
         cmpw    $ASK_VGA, %ax                   # Bring up the menu
@@ -69,7 +50,7 @@ vid1:   call    store_edid
 
 # Detect if we have CGA, MDA, EGA or VGA and pass it to the kernel.
 basic_detect:
-        movb    $0, _param(PARAM_HAVE_VGA)
+        movb    $0, _param(BVI_have_vga)
         movb    $0x12, %ah                      # Check EGA/VGA
         movb    $0x10, %bl
         int     $0x10
@@ -79,7 +60,7 @@ basic_detect:
         int     $0x10
         cmpb    $0x1a, %al                      # 1a means VGA...
         jne     basret                          # anything else is EGA.
-        incb    _param(PARAM_HAVE_VGA)          # We've detected a VGA
+        incb    _param(BVI_have_vga)            # We've detected a VGA
 basret: ret
 
 # Store the video mode parameters for later usage by the kernel.
@@ -92,57 +73,57 @@ mode_params:
         movb    $0x03, %ah                      # Read cursor position
         xorb    %bh, %bh
         int     $0x10
-        movw    %dx, _param(PARAM_CURSOR_POS)
+        movw    %dx, _param(BVI_cursor_pos)
         movb    $0x0f, %ah                      # Read page/mode/width
         int     $0x10
-        movw    %ax, _param(PARAM_VIDEO_MODE)   # Video mode and screen width
+        movw    %ax, _param(BVI_video_mode)     # Video mode and screen width
         movw    %gs:(0x485), %ax                # Font size
-        movw    %ax, _param(PARAM_FONT_POINTS)  # (valid only on EGA/VGA)
+        movw    %ax, _param(BVI_font_points)    # (valid only on EGA/VGA)
         movw    bootsym(force_size), %ax        # Forced size?
         orw     %ax, %ax
         jz      mopar1
 
-        movb    %ah, _param(PARAM_VIDEO_COLS)
-        movb    %al, _param(PARAM_VIDEO_LINES)
+        movb    %ah, _param(BVI_video_cols)
+        movb    %al, _param(BVI_video_lines)
         ret
 
 mopar1: movb    %gs:(0x484), %al                # On EGA/VGA, use the EGA+ BIOS
         incb    %al                             # location of max lines.
-mopar2: movb    %al, _param(PARAM_VIDEO_LINES)
+mopar2: movb    %al, _param(BVI_video_lines)
         ret
 
 # Fetching of VESA frame buffer parameters
 mopar_gr:
         movw    $vesa_mode_info, %di
-        movb    $0x23, _param(PARAM_HAVE_VGA)
+        movb    $0x23, _param(BVI_have_vga)
         movw    16(%di), %ax
-        movw    %ax, _param(PARAM_LFB_LINELENGTH)
+        movw    %ax, _param(BVI_lfb_linelength)
         movw    18(%di), %ax
-        movw    %ax, _param(PARAM_LFB_WIDTH)
+        movw    %ax, _param(BVI_lfb_width)
         movw    20(%di), %ax
-        movw    %ax, _param(PARAM_LFB_HEIGHT)
+        movw    %ax, _param(BVI_lfb_height)
         movzbw  25(%di), %ax
-        movw    %ax, _param(PARAM_LFB_DEPTH)
+        movw    %ax, _param(BVI_lfb_depth)
         movl    40(%di), %eax
-        movl    %eax, _param(PARAM_LFB_BASE)
+        movl    %eax, _param(BVI_lfb_base)
         movl    31(%di), %eax
-        movl    %eax, _param(PARAM_LFB_COLORS)
+        movl    %eax, _param(BVI_lfb_colors)
         movl    35(%di), %eax
-        movl    %eax, _param(PARAM_LFB_COLORS+4)
+        movl    %eax, _param(BVI_lfb_colors+4)
         movw    0(%di), %ax
-        movw    %ax, _param(PARAM_VESA_ATTRIB)
+        movw    %ax, _param(BVI_vesa_attrib)
 
 # get video mem size
         movw    $vesa_glob_info, %di
         movzwl  18(%di), %eax
-        movl    %eax, _param(PARAM_LFB_SIZE)
+        movl    %eax, _param(BVI_lfb_size)
 
 # store mode capabilities
         movl    10(%di), %eax
-        movl    %eax, _param(PARAM_CAPABILITIES)
+        movl    %eax, _param(BVI_capabilities)
 
 # switching the DAC to 8-bit is for <= 8 bpp only
-        cmpw    $8, _param(PARAM_LFB_DEPTH)
+        cmpw    $8, _param(BVI_lfb_depth)
         jg      dac_done
 
 # get DAC switching capability
@@ -160,16 +141,16 @@ mopar_gr:
 dac_set:
 # set color size to DAC size
         movzbw  bootsym(dac_size), %ax
-        movb    %al, _param(PARAM_LFB_COLORS+0)
-        movb    %al, _param(PARAM_LFB_COLORS+2)
-        movb    %al, _param(PARAM_LFB_COLORS+4)
-        movb    %al, _param(PARAM_LFB_COLORS+6)
+        movb    %al, _param(BVI_lfb_colors+0)
+        movb    %al, _param(BVI_lfb_colors+2)
+        movb    %al, _param(BVI_lfb_colors+4)
+        movb    %al, _param(BVI_lfb_colors+6)
 
 # set color offsets to 0
-        movb    %ah, _param(PARAM_LFB_COLORS+1)
-        movb    %ah, _param(PARAM_LFB_COLORS+3)
-        movb    %ah, _param(PARAM_LFB_COLORS+5)
-        movb    %ah, _param(PARAM_LFB_COLORS+7)
+        movb    %ah, _param(BVI_lfb_colors+1)
+        movb    %ah, _param(BVI_lfb_colors+3)
+        movb    %ah, _param(BVI_lfb_colors+5)
+        movb    %ah, _param(BVI_lfb_colors+7)
 
 dac_done:
 # get protected mode interface information
@@ -180,8 +161,8 @@ dac_done:
         cmp     $0x004f, %ax
         jnz     no_pm
 
-        movw    %es, _param(PARAM_VESAPM_SEG)
-        movw    %di, _param(PARAM_VESAPM_OFF)
+        movw    %es, _param(BVI_vesapm_seg)
+        movw    %di, _param(BVI_vesapm_off)
 
 no_pm:  pushw   %ds
         popw    %es
@@ -1018,7 +999,7 @@ GLOBAL(boot_vid_info)
         .byte   80, 25  /* 80x25          */
         .byte   1       /* isVGA          */
         .word   16      /* 8x16 font      */
-        .fill   0x28,1,0
+        .space  BVI_size - BVI_capabilities
 GLOBAL(boot_edid_info)
         .fill   128,1,0x13
 GLOBAL(boot_edid_caps)
diff --git a/xen/arch/x86/x86_64/asm-offsets.c b/xen/arch/x86/x86_64/asm-offsets.c
index d8903a3ce9c7..91da6b9d3885 100644
--- a/xen/arch/x86/x86_64/asm-offsets.c
+++ b/xen/arch/x86/x86_64/asm-offsets.c
@@ -16,6 +16,10 @@
 #include <xen/multiboot.h>
 #include <xen/multiboot2.h>
 
+#ifdef CONFIG_VIDEO
+# include "../boot/video.h"
+#endif
+
 #define DEFINE(_sym, _val)                                                 \
     asm volatile ( "\n.ascii\"==>#define " #_sym " %0 /* " #_val " */<==\""\
                    :: "i" (_val) )
@@ -208,4 +212,26 @@ void __dummy__(void)
 
     OFFSET(DOMAIN_vm_assist, struct domain, vm_assist);
     BLANK();
+
+#ifdef CONFIG_VIDEO
+    DEFINE(BVI_size, sizeof(struct boot_video_info));
+    OFFSET(BVI_cursor_pos, struct boot_video_info, orig_x);
+    OFFSET(BVI_video_mode, struct boot_video_info, orig_video_mode);
+    OFFSET(BVI_video_cols, struct boot_video_info, orig_video_cols);
+    OFFSET(BVI_video_lines, struct boot_video_info, orig_video_lines);
+    OFFSET(BVI_have_vga, struct boot_video_info, orig_video_isVGA);
+    OFFSET(BVI_font_points, struct boot_video_info, orig_video_points);
+    OFFSET(BVI_capabilities, struct boot_video_info, capabilities);
+    OFFSET(BVI_lfb_linelength, struct boot_video_info, lfb_linelength);
+    OFFSET(BVI_lfb_width, struct boot_video_info, lfb_width);
+    OFFSET(BVI_lfb_height, struct boot_video_info, lfb_height);
+    OFFSET(BVI_lfb_depth, struct boot_video_info, lfb_depth);
+    OFFSET(BVI_lfb_base, struct boot_video_info, lfb_base);
+    OFFSET(BVI_lfb_size, struct boot_video_info, lfb_size);
+    OFFSET(BVI_lfb_colors, struct boot_video_info, colors);
+    OFFSET(BVI_vesapm_seg, struct boot_video_info, vesapm.seg);
+    OFFSET(BVI_vesapm_off, struct boot_video_info, vesapm.off);
+    OFFSET(BVI_vesa_attrib, struct boot_video_info, vesa_attrib);
+    BLANK();
+#endif /* CONFIG_VIDEO */
 }
-- 
2.44.0

[PATCH 2/2] x86/video: do not assume a video mode to be unconditionally present

[Roger Pau Monne]

There's no reason to assume VGA text mode 3 to be unconditionally available.
With the addition of booting Xen itself in PVH mode there's a boot path that
explicitly short-circuits all the real-mode logic, including the VGA detection.

Leave the default user selected mode as text mode 3 in boot_vid_mode, but do
not populate boot_vid_info with any default settings.  It will either be
populated by the real-mode video detection code, or left zeroed in case
real-mode code is skipped.

Note that only PVH skips the real-mode portion of the boot trampoline,
otherwise the only way to skip it is to set `no-real-mode` on the command line,
and the description for the option already notes that VGA would be disabled as
a result of skipping real-mode bootstrap.

This fixes Xen incorrectly reporting:

(XEN) Video information:
(XEN)  VGA is text mode 80x25, font 8x16

When booted as a PVH guest.

Reported-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
---
 xen/arch/x86/boot/video.S | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/xen/arch/x86/boot/video.S b/xen/arch/x86/boot/video.S
index a4b25a3b34d1..a51de04a024e 100644
--- a/xen/arch/x86/boot/video.S
+++ b/xen/arch/x86/boot/video.S
@@ -994,12 +994,7 @@ name_bann:      .asciz  "Video adapter: "
 force_size:     .word   0       # Use this size instead of the one in BIOS vars
 
 GLOBAL(boot_vid_info)
-        .byte   0, 0    /* orig_x, orig_y */
-        .byte   3       /* text mode 3    */
-        .byte   80, 25  /* 80x25          */
-        .byte   1       /* isVGA          */
-        .word   16      /* 8x16 font      */
-        .space  BVI_size - BVI_capabilities
+        .space  BVI_size
 GLOBAL(boot_edid_info)
         .fill   128,1,0x13
 GLOBAL(boot_edid_caps)
-- 
2.44.0

Re: [PATCH 1/2] x86/video: add boot_video_info offset generation to asm-offsets

[Andrew Cooper]

On 28/03/2024 3:35 pm, Roger Pau Monne wrote:
> Currently the offsets into the boot_video_info struct are manually encoded in
> video.S, which is fragile.  Generate them in asm-offsets.c and switch the
> current code to use those instead.
>
> No functional change intended.
>
> Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>

R-by: Andrew Cooper <andrew.cooper3@citrix.com>

Thanks for looking at this.  As you're touching these lines, there are a
few style fixes.

> diff --git a/xen/arch/x86/boot/video.S b/xen/arch/x86/boot/video.S
> index 0ae04f270f8c..a4b25a3b34d1 100644
> --- a/xen/arch/x86/boot/video.S
> +++ b/xen/arch/x86/boot/video.S
> @@ -26,32 +26,13 @@
>  /* Force 400 scan lines for standard modes (hack to fix bad BIOS behaviour */
>  #undef CONFIG_VIDEO_400_HACK
>  
> -/* Positions of various video parameters passed to the kernel */
> -/* (see also include/linux/tty.h) */
> -#define PARAM_CURSOR_POS        0x00
> -#define PARAM_VIDEO_MODE        0x02
> -#define PARAM_VIDEO_COLS        0x03
> -#define PARAM_VIDEO_LINES       0x04
> -#define PARAM_HAVE_VGA          0x05
> -#define PARAM_FONT_POINTS       0x06
> -#define PARAM_CAPABILITIES      0x08
> -#define PARAM_LFB_LINELENGTH    0x0c
> -#define PARAM_LFB_WIDTH         0x0e
> -#define PARAM_LFB_HEIGHT        0x10
> -#define PARAM_LFB_DEPTH         0x12
> -#define PARAM_LFB_BASE          0x14
> -#define PARAM_LFB_SIZE          0x18
> -#define PARAM_LFB_COLORS        0x1c
> -#define PARAM_VESAPM_SEG        0x24
> -#define PARAM_VESAPM_OFF        0x26
> -#define PARAM_VESA_ATTRIB       0x28
>  #define _param(param) bootsym(boot_vid_info)+(param)
>  
>  video:  xorw    %ax, %ax
>          movw    %ax, %gs        # GS is zero
>          cld
>          call    basic_detect    # Basic adapter type testing (EGA/VGA/MDA/CGA)
> -        cmpb    $0,_param(PARAM_HAVE_VGA)
> +        cmpb    $0,_param(BVI_have_vga)

Space

> @@ -160,16 +141,16 @@ mopar_gr:
>  dac_set:
>  # set color size to DAC size
>          movzbw  bootsym(dac_size), %ax
> -        movb    %al, _param(PARAM_LFB_COLORS+0)
> -        movb    %al, _param(PARAM_LFB_COLORS+2)
> -        movb    %al, _param(PARAM_LFB_COLORS+4)
> -        movb    %al, _param(PARAM_LFB_COLORS+6)
> +        movb    %al, _param(BVI_lfb_colors+0)
> +        movb    %al, _param(BVI_lfb_colors+2)
> +        movb    %al, _param(BVI_lfb_colors+4)
> +        movb    %al, _param(BVI_lfb_colors+6)

Spaces

> diff --git a/xen/arch/x86/x86_64/asm-offsets.c b/xen/arch/x86/x86_64/asm-offsets.c
> index d8903a3ce9c7..91da6b9d3885 100644
> --- a/xen/arch/x86/x86_64/asm-offsets.c
> +++ b/xen/arch/x86/x86_64/asm-offsets.c
> @@ -16,6 +16,10 @@
>  #include <xen/multiboot.h>
>  #include <xen/multiboot2.h>
>  
> +#ifdef CONFIG_VIDEO
> +# include "../boot/video.h"
> +#endif
> +
>  #define DEFINE(_sym, _val)                                                 \
>      asm volatile ( "\n.ascii\"==>#define " #_sym " %0 /* " #_val " */<==\""\
>                     :: "i" (_val) )
> @@ -208,4 +212,26 @@ void __dummy__(void)
>  
>      OFFSET(DOMAIN_vm_assist, struct domain, vm_assist);
>      BLANK();
> +
> +#ifdef CONFIG_VIDEO
> +    DEFINE(BVI_size, sizeof(struct boot_video_info));
> +    OFFSET(BVI_cursor_pos, struct boot_video_info, orig_x);
> +    OFFSET(BVI_video_mode, struct boot_video_info, orig_video_mode);
> +    OFFSET(BVI_video_cols, struct boot_video_info, orig_video_cols);
> +    OFFSET(BVI_video_lines, struct boot_video_info, orig_video_lines);
> +    OFFSET(BVI_have_vga, struct boot_video_info, orig_video_isVGA);
> +    OFFSET(BVI_font_points, struct boot_video_info, orig_video_points);
> +    OFFSET(BVI_capabilities, struct boot_video_info, capabilities);
> +    OFFSET(BVI_lfb_linelength, struct boot_video_info, lfb_linelength);
> +    OFFSET(BVI_lfb_width, struct boot_video_info, lfb_width);
> +    OFFSET(BVI_lfb_height, struct boot_video_info, lfb_height);
> +    OFFSET(BVI_lfb_depth, struct boot_video_info, lfb_depth);
> +    OFFSET(BVI_lfb_base, struct boot_video_info, lfb_base);
> +    OFFSET(BVI_lfb_size, struct boot_video_info, lfb_size);
> +    OFFSET(BVI_lfb_colors, struct boot_video_info, colors);
> +    OFFSET(BVI_vesapm_seg, struct boot_video_info, vesapm.seg);
> +    OFFSET(BVI_vesapm_off, struct boot_video_info, vesapm.off);
> +    OFFSET(BVI_vesa_attrib, struct boot_video_info, vesa_attrib);
> +    BLANK();
> +#endif /* CONFIG_VIDEO */

BVI_size traditionally goes last.  MB2 (which I guess you copied?) is a
little odd.

I'd like to start vertically aligning this for readability, like I
started with EFRAME_*.

Happy to fold of all of these tweaks on commit.

~Andrew

Re: [PATCH 2/2] x86/video: do not assume a video mode to be unconditionally present

[Andrew Cooper]

On 28/03/2024 3:35 pm, Roger Pau Monne wrote:
> There's no reason to assume VGA text mode 3 to be unconditionally available.
> With the addition of booting Xen itself in PVH mode there's a boot path that
> explicitly short-circuits all the real-mode logic, including the VGA detection.
>
> Leave the default user selected mode as text mode 3 in boot_vid_mode, but do
> not populate boot_vid_info with any default settings.  It will either be
> populated by the real-mode video detection code, or left zeroed in case
> real-mode code is skipped.
>
> Note that only PVH skips the real-mode portion of the boot trampoline,
> otherwise the only way to skip it is to set `no-real-mode` on the command line,
> and the description for the option already notes that VGA would be disabled as
> a result of skipping real-mode bootstrap.

IIRC, Grub defaults to using no-real-mode for xen.efi.  It's definitely
a common option to find used in practice.

>
> This fixes Xen incorrectly reporting:
>
> (XEN) Video information:
> (XEN)  VGA is text mode 80x25, font 8x16
>
> When booted as a PVH guest.
>
> Reported-by: Andrew Cooper <andrew.cooper3@citrix.com>
> Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>

Reivewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
Tested-by: Andrew Cooper <andrew.cooper3@citrix.com>

Re: [PATCH 1/2] x86/video: add boot_video_info offset generation to asm-offsets

[Jan Beulich]

On 28.03.2024 16:35, Roger Pau Monne wrote:
> Currently the offsets into the boot_video_info struct are manually encoded in
> video.S, which is fragile.  Generate them in asm-offsets.c and switch the
> current code to use those instead.

Just to mention it (without asking for immediate action): Defining boot_vid_info
in assembly code then is as fragile. Moving to C would likely be problematic
because it needs to be in the trampoline range. But at least its size should (at
some point) perhaps better be tied to the C struct's sizeof(). The fields, with
some effort, could also be converted using the new BVI_* constants. That would
still leave the field sizes; maybe those could at least be cross-checked by some
BUILD_BUG_ONs.

Jan

Re: [PATCH 1/2] x86/video: add boot_video_info offset generation to asm-offsets

[Jan Beulich]

On 02.04.2024 11:38, Jan Beulich wrote:
> On 28.03.2024 16:35, Roger Pau Monne wrote:
>> Currently the offsets into the boot_video_info struct are manually encoded in
>> video.S, which is fragile.  Generate them in asm-offsets.c and switch the
>> current code to use those instead.
> 
> Just to mention it (without asking for immediate action): Defining boot_vid_info
> in assembly code then is as fragile. Moving to C would likely be problematic
> because it needs to be in the trampoline range. But at least its size should (at
> some point) perhaps better be tied to the C struct's sizeof().

Actually I overlooked that you partly do this. The use of BVI_capabilities there
looks odd to me, though. Why not

        .space  BVI_size - (. - boot_vid_info)

? I realize it becomes just BVI_size in patch 2, but I have some question there,
too.

Jan

> The fields, with
> some effort, could also be converted using the new BVI_* constants. That would
> still leave the field sizes; maybe those could at least be cross-checked by some
> BUILD_BUG_ONs.
> 
> Jan

Re: [PATCH 2/2] x86/video: do not assume a video mode to be unconditionally present

[Jan Beulich]

On 28.03.2024 16:35, Roger Pau Monne wrote:
> There's no reason to assume VGA text mode 3 to be unconditionally available.
> With the addition of booting Xen itself in PVH mode there's a boot path that
> explicitly short-circuits all the real-mode logic, including the VGA detection.
> 
> Leave the default user selected mode as text mode 3 in boot_vid_mode, but do
> not populate boot_vid_info with any default settings.  It will either be
> populated by the real-mode video detection code, or left zeroed in case
> real-mode code is skipped.
> 
> Note that only PVH skips the real-mode portion of the boot trampoline,
> otherwise the only way to skip it is to set `no-real-mode` on the command line,
> and the description for the option already notes that VGA would be disabled as
> a result of skipping real-mode bootstrap.
> 
> This fixes Xen incorrectly reporting:
> 
> (XEN) Video information:
> (XEN)  VGA is text mode 80x25, font 8x16
> 
> When booted as a PVH guest.

And what effect does this have on a bare-metal boot with no-real-mode in use?
The default on x86 hardware still is that in the absence of other information,
a VGA of some kind can be assumed to be there. Yes, there are headless
systems, but better assume VGA is there when there's not than the other way
around.

What I would have expected is for the PVH boot path to clear boot_vid_info.

Jan

Re: [PATCH 2/2] x86/video: do not assume a video mode to be unconditionally present

[Jan Beulich]

On 28.03.2024 20:58, Andrew Cooper wrote:
> On 28/03/2024 3:35 pm, Roger Pau Monne wrote:
>> There's no reason to assume VGA text mode 3 to be unconditionally available.
>> With the addition of booting Xen itself in PVH mode there's a boot path that
>> explicitly short-circuits all the real-mode logic, including the VGA detection.
>>
>> Leave the default user selected mode as text mode 3 in boot_vid_mode, but do
>> not populate boot_vid_info with any default settings.  It will either be
>> populated by the real-mode video detection code, or left zeroed in case
>> real-mode code is skipped.
>>
>> Note that only PVH skips the real-mode portion of the boot trampoline,
>> otherwise the only way to skip it is to set `no-real-mode` on the command line,
>> and the description for the option already notes that VGA would be disabled as
>> a result of skipping real-mode bootstrap.
> 
> IIRC, Grub defaults to using no-real-mode for xen.efi.

That's our MB2 entry path which forces skip_realmode to 1. The xen.efi boot
path doesn't, but similarly skips entering real mode (retrieving desired
data via EFI protocols instead). Imo if to be retained, the above paragraph
would want extending some, to cover all the cases.

Jan

Re: [PATCH 2/2] x86/video: do not assume a video mode to be unconditionally present

[Andrew Cooper]

On 02/04/2024 10:54 am, Jan Beulich wrote:
> On 28.03.2024 20:58, Andrew Cooper wrote:
>> On 28/03/2024 3:35 pm, Roger Pau Monne wrote:
>>> There's no reason to assume VGA text mode 3 to be unconditionally available.
>>> With the addition of booting Xen itself in PVH mode there's a boot path that
>>> explicitly short-circuits all the real-mode logic, including the VGA detection.
>>>
>>> Leave the default user selected mode as text mode 3 in boot_vid_mode, but do
>>> not populate boot_vid_info with any default settings.  It will either be
>>> populated by the real-mode video detection code, or left zeroed in case
>>> real-mode code is skipped.
>>>
>>> Note that only PVH skips the real-mode portion of the boot trampoline,
>>> otherwise the only way to skip it is to set `no-real-mode` on the command line,
>>> and the description for the option already notes that VGA would be disabled as
>>> a result of skipping real-mode bootstrap.
>> IIRC, Grub defaults to using no-real-mode for xen.efi.
> That's our MB2 entry path which forces skip_realmode to 1. The xen.efi boot
> path doesn't, but similarly skips entering real mode (retrieving desired
> data via EFI protocols instead). Imo if to be retained, the above paragraph
> would want extending some, to cover all the cases.

What I mean is that Grub's 20_linux_xen script writes a stanza which
includes "no-real-mode edd=off" for any non-PC platform, which includes EFI.

~Andrew

Re: [PATCH 1/2] x86/video: add boot_video_info offset generation to asm-offsets

[Roger Pau Monné]

On Tue, Apr 02, 2024 at 11:43:49AM +0200, Jan Beulich wrote:
> On 02.04.2024 11:38, Jan Beulich wrote:
> > On 28.03.2024 16:35, Roger Pau Monne wrote:
> >> Currently the offsets into the boot_video_info struct are manually encoded in
> >> video.S, which is fragile.  Generate them in asm-offsets.c and switch the
> >> current code to use those instead.
> > 
> > Just to mention it (without asking for immediate action): Defining boot_vid_info
> > in assembly code then is as fragile. Moving to C would likely be problematic
> > because it needs to be in the trampoline range. But at least its size should (at
> > some point) perhaps better be tied to the C struct's sizeof().
> 
> Actually I overlooked that you partly do this. The use of BVI_capabilities there
> looks odd to me, though. Why not
> 
>         .space  BVI_size - (. - boot_vid_info)
> 
> ? I realize it becomes just BVI_size in patch 2, but I have some question there,
> too.

I didn't think much about this because it was going away in the next
patch, but let me adjust.  In fact I was tempted to just use BVI_size
and allocate more than required.

Thanks, Roger.

Re: [PATCH 2/2] x86/video: do not assume a video mode to be unconditionally present

[Roger Pau Monné]

On Tue, Apr 02, 2024 at 11:49:20AM +0200, Jan Beulich wrote:
> On 28.03.2024 16:35, Roger Pau Monne wrote:
> > There's no reason to assume VGA text mode 3 to be unconditionally available.
> > With the addition of booting Xen itself in PVH mode there's a boot path that
> > explicitly short-circuits all the real-mode logic, including the VGA detection.
> > 
> > Leave the default user selected mode as text mode 3 in boot_vid_mode, but do
> > not populate boot_vid_info with any default settings.  It will either be
> > populated by the real-mode video detection code, or left zeroed in case
> > real-mode code is skipped.
> > 
> > Note that only PVH skips the real-mode portion of the boot trampoline,
> > otherwise the only way to skip it is to set `no-real-mode` on the command line,
> > and the description for the option already notes that VGA would be disabled as
> > a result of skipping real-mode bootstrap.
> > 
> > This fixes Xen incorrectly reporting:
> > 
> > (XEN) Video information:
> > (XEN)  VGA is text mode 80x25, font 8x16
> > 
> > When booted as a PVH guest.
> 
> And what effect does this have on a bare-metal boot with no-real-mode in use?
> The default on x86 hardware still is that in the absence of other information,
> a VGA of some kind can be assumed to be there. Yes, there are headless
> systems, but better assume VGA is there when there's not than the other way
> around.

But that contradicts the text of the 'no-real-mode' option, which
explicitly notes:

"Do not execute real-mode bootstrap code when booting Xen. This option
should not be used except for debugging. It will effectively disable
the vga option, which relies on real mode to set the video mode."

> What I would have expected is for the PVH boot path to clear boot_vid_info.

Well, my intention was to fix both PVH and also make the
implementation of the 'no-real-mode' option consistent with the
documentation.

Thanks, Roger.

Re: [PATCH 2/2] x86/video: do not assume a video mode to be unconditionally present

[Jan Beulich]

On 19.04.2024 09:42, Roger Pau Monné wrote:
> On Tue, Apr 02, 2024 at 11:49:20AM +0200, Jan Beulich wrote:
>> On 28.03.2024 16:35, Roger Pau Monne wrote:
>>> There's no reason to assume VGA text mode 3 to be unconditionally available.
>>> With the addition of booting Xen itself in PVH mode there's a boot path that
>>> explicitly short-circuits all the real-mode logic, including the VGA detection.
>>>
>>> Leave the default user selected mode as text mode 3 in boot_vid_mode, but do
>>> not populate boot_vid_info with any default settings.  It will either be
>>> populated by the real-mode video detection code, or left zeroed in case
>>> real-mode code is skipped.
>>>
>>> Note that only PVH skips the real-mode portion of the boot trampoline,
>>> otherwise the only way to skip it is to set `no-real-mode` on the command line,
>>> and the description for the option already notes that VGA would be disabled as
>>> a result of skipping real-mode bootstrap.
>>>
>>> This fixes Xen incorrectly reporting:
>>>
>>> (XEN) Video information:
>>> (XEN)  VGA is text mode 80x25, font 8x16
>>>
>>> When booted as a PVH guest.
>>
>> And what effect does this have on a bare-metal boot with no-real-mode in use?
>> The default on x86 hardware still is that in the absence of other information,
>> a VGA of some kind can be assumed to be there. Yes, there are headless
>> systems, but better assume VGA is there when there's not than the other way
>> around.
> 
> But that contradicts the text of the 'no-real-mode' option, which
> explicitly notes:
> 
> "Do not execute real-mode bootstrap code when booting Xen. This option
> should not be used except for debugging. It will effectively disable
> the vga option, which relies on real mode to set the video mode."

Well. Even without setting a video mode, _some_ mode is set (by firmware) as
long as there is a VGA. In the absence of a "vga=" option iirc we'd retrieve
that setting, unless we're not allowed to by "no-real-mode". In which case,
as indicated, we may still be better off guessing a basic mode than kind of
suggesting the absence of any VGA.

Besides, when booting from EFI vga= is respected despite not going through
real mode. So the quoted text isn't quite right anyway.

Jan

>> What I would have expected is for the PVH boot path to clear boot_vid_info.
> 
> Well, my intention was to fix both PVH and also make the
> implementation of the 'no-real-mode' option consistent with the
> documentation.
> 
> Thanks, Roger.