Kevin O'Connor [Sat, 18 Feb 2012 16:02:27 +0000 (11:02 -0500)]
Register drives directly in scsi_init_drive().
The scsi_init_drive() function has enough information to directly
register the drive there, so do the registration there. This
simplifies the calling code.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Gleb Natapov [Mon, 13 Feb 2012 11:01:07 +0000 (13:01 +0200)]
Running vgabios during resume from S3 on QEMU by default
Run vgabios during resume from S3 by default on QEMU. QEMU
still able to modify SeaBIOS behavior if it wishes so by providing
etc/s3-resume-vga-init file. With QEMU emulated vga cards this behaviour
is desirable otherwise console becomes unusable with Linux guests after
resume. Since we control vgabios source we can be sure that running it
on resume from S3 is safe.
Kevin O'Connor [Tue, 14 Feb 2012 01:09:02 +0000 (20:09 -0500)]
vgabios: Some tweaks to optimize stack space.
Gcc seems to handle passing structs by value if they are 4 bytes in
size instead of 3 bytes. So, add a pad byte to struct carattr and
struct cursorpos.
Reorganize set_cursor_pos(), verify_scroll(), and handle_1013() so
there are less live variables.
Don't inline the VBE functions into the main handler code. Code
calling VBE functions are newer and are more likely to provide
adequate stack space - inlining can cause more stack usage for older
functions (which may be stack constrained).
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sat, 11 Feb 2012 15:49:45 +0000 (10:49 -0500)]
Direct compile 16bit C code instead of including via .S files.
Create a ".code16gcc" directive in src/code16gcc.s and use
-Wa,src/code16gcc.s to tell gcc to compile directly to 16bit code.
This eliminates the need to compile the C code to assembler and
include in romlayout.S and vgaentry.S. This also allows those two
assembler files to be compiled with debugging (-g) enabled.
Also, includes some Makefile cleanups.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sat, 11 Feb 2012 14:38:44 +0000 (09:38 -0500)]
disk: handle LBA I/O with zero sector count
Unlike basic_access, extended_access does not check for a zero
sector count. However, this is a problem because for example
it would be interpreted as 256 when processing an ATA request.
Based on patch from: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Commit 1e749c85 removed the definition of process_usb_op(); let's remove
the declaration too.
The same commit added process_scsi_op(). The function has no declaration
that is also not a definition, and its only call site is in the same file
as the definition. Give the function internal linkage.
Julian Pidancet [Sun, 5 Feb 2012 04:51:06 +0000 (04:51 +0000)]
Xen: Use VGA Hooks to make VGA passthrough work on certain devices
The Intel gfx VGA option ROM on certain platforms requires the 155f50 BIOS
function to be implemented (even if it does nothing), to work properly.
v2: Ignore the case where Option ROMs are pre-deployed.
Make vgahook_setup independent of wether the CB* variables are set.
VGA hooks can be enabled on non-coreboot and non-Xen configurations.
Kevin O'Connor [Fri, 3 Feb 2012 03:52:17 +0000 (22:52 -0500)]
vgabios: Handle VGA option rom being re-run.
Check for the case where the option rom is executed a second time - in
that case re-init the hardware, but do not set any variables. This
should make the rom better behaving when run from S3 resume.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Gleb Natapov [Tue, 24 Jan 2012 12:33:42 +0000 (14:33 +0200)]
mask interrupts on S3 resume
i8259 clears interrupt mask on reset. Interrupt need to be
masked again before enabling interrupts on CPU. Since option
roms are called with interrupts enabled, resume should mask
interrupts in i8259 before calling vgabios.
Signed-off-by: Gleb Natapov <gleb@redhat.com> Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sun, 29 Jan 2012 16:53:59 +0000 (11:53 -0500)]
vgabios: handle vmware vga in bochsvga.
The vmware vga emulated by qemu has a I/O region in pci bar 0.
The framebuffer is in pci bar 1. Handle that by checking the
type of bar 0 in case it is a I/O bar use bar 1 instead.
Also make bochsbios report lfb size in debug output.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com> Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Mon, 16 Jan 2012 23:48:26 +0000 (18:48 -0500)]
vgabios: Simplify planar4 vgafb code.
Don't bother programming the vga registers to manipulate the
framebuffer when in planar4 mode. Instead, just switch between the
four planes and do the manipulation with regular reads and writes.
This makes the code simpler to understand (the vga hardware
manipulations are arcane and complex).
Note, this could make text scrolling in planar4 mode slower and more
likely to result in tearing. However, it's unlikely anything
important uses the vgabios in planar4 mode.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sun, 15 Jan 2012 06:12:20 +0000 (01:12 -0500)]
vgabios: Improve vgabios Kconfig menu.
Place driver types in a "choice" selection - only one driver type
makes sense.
Only prompt the user for PCI Vendor/Device ids if they wish to
override the default choices. Otherwise, Kconfig wont know to use the
proper defaults when building autoconf.h.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sat, 14 Jan 2012 19:52:01 +0000 (14:52 -0500)]
vgabios: Make cirrus line lengths standard.
Only two modes don't match their expected line length. One looks like
a bug (it has a virtual line length of 1280 for a screen of 1600
bytes) and one looks like an optimization (2048 vs 1920). Change the
mode line lengths to exactly match the expected line lengths so that
the VBE code is correct.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sat, 14 Jan 2012 01:00:35 +0000 (20:00 -0500)]
vgabios: Make VBE code independent of bochsvga.
Introduce new global variables (VBE_enabled, VBE_total_memory,
VBE_capabilities, VBE_framebuffer) to replace the need for function
calls that were specific to bochsvga.
Replace info received from bochsvga_mode_info with info found in
vgahw_find_mode.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sat, 7 Jan 2012 23:27:19 +0000 (18:27 -0500)]
vgabios: Unify page size calculations; remove page size from vgamode_s.
The previous code could obtain the page size in different ways - from
vmode_g->sslength, from SCREEN_MEM_START, or by manually multiplying
cols and rows. Add a new func (calc_page_size) and use that in areas
that need to calculate the page size.
Convert readers of the page size to read it from the "video_pagesize"
entry in the BDA instead of recalculating it.
Remove the page size from struct vgamode_s (slength) as it is now
calculated dynamically. The new calculated versions are different
from the existing values exported in video_param_table. However, the
existing values (for CGA) did not look correct - I compared the values
to those exported by two other VGA BIOS manufacturers and used the
values that look more standard.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Mon, 2 Jan 2012 16:13:14 +0000 (11:13 -0500)]
vgabios: Make struct vgamode_s more similar to bochs/cirrus mode tables.
For graphics modes, store pixel width/height instead of text
width/height. Add explicit char width field. Where needed, calculate
text width/height from pixel width/height by dividing pixel count by
character size.
Rename some fields and change field sizes to match cirrus/bochs
definitions.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sat, 14 Jan 2012 16:58:14 +0000 (11:58 -0500)]
BCVs should inherrit the legacy harddrive priority.
A BCV will (almost assuredly) be a hard drive, so it makes sense to
prioritize them the same as builtin hard drives when only the legacy
priority system is used.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sat, 14 Jan 2012 16:55:35 +0000 (11:55 -0500)]
Permit multiple BEV and BEV/BCV combinations in a single option rom.
The BBS spec is unclear on multiple BEV entry points for a rom.
However, the LSI scsi rom has been seen to register a BEV followed by
multiple BCV entry points. Add support for it, as there's no harm in
it.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sun, 1 Jan 2012 15:54:19 +0000 (10:54 -0500)]
vgabios: Minor - make FB scroll functions look similar.
All three scroll function variants have the same format, but those
that have an implicit character width or height of 1 have those
multiplications omitted. Add those multiplications back in (the
compiler will optimize them away) so all the variants look similar.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sat, 31 Dec 2011 23:19:22 +0000 (18:19 -0500)]
vgabios: Refactor vga_set_mode and stdvga_set_mode.
Split out the BDA setup part of vga_set_mode to new function
modeswitch_set_bda. Move the remaining parts (palette loading, screen
clearing, font loading) of vga_set_mode into stdvga_set_mode.
Add new mode switching flags and pass them to stdvga_set_mode, so it
does not need to inspect modeset_ctl directly.
Move code needed by stdvga_set_mode (perform_gray_scale_summing,
clear_screen) to stdvga.c.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sat, 31 Dec 2011 22:24:11 +0000 (17:24 -0500)]
vgabios: Change vga_set_mode() to directly setup BDA.
The calls to set_cursor_shape, set_cursor_pos, and set_active_page
have no impact other than to set the BDA variables, because the
standard vga mode switch already programed the registers these
functions set. So, just setup the BDA directly.
The stdvga_set_text_block_specifier call is unnecessary - it sets
a register which is already programmed that way.
Call stdvga_get_crtc() to obtain the crtc address instead of inferring
it from the memory model.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sat, 31 Dec 2011 21:22:35 +0000 (16:22 -0500)]
vgabios: Don't call int10 during mode switch.
Don't call int10 to load the fonts - instead call the font loading
functions directly.
Theoretically, this could change behavior if an external program has
captured int10 and redirects the font loading calls. However, there
does not seem to be any indication that an external app could expect
that to work.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
projects / seabios.git / log