openfirmware February 2009

openfirmware@openfirmware.info

3 participants
21 discussions

by svn＠openfirmware.info

Author: wmb Date: 2009-02-20 22:50:02 +0100 (Fri, 20 Feb 2009) New Revision: 1115 Modified: cpu/x86/pc/olpc/versions.fth Log: Q2E33 Modified: cpu/x86/pc/olpc/versions.fth =================================================================== --- cpu/x86/pc/olpc/versions.fth 2009-02-20 21:44:09 UTC (rev 1114) +++ cpu/x86/pc/olpc/versions.fth 2009-02-20 21:50:02 UTC (rev 1115) @@ -2,7 +2,7 @@ \ The overall firmware revision macro: FW_MAJOR E -macro: FW_MINOR 32 +macro: FW_MINOR 33 \ The EC microcode macro: EC_VERSION e32

11 years, 5 months

r1114 - cpu/x86/pc/olpc

by svn＠openfirmware.info

Author: wmb Date: 2009-02-20 22:44:09 +0100 (Fri, 20 Feb 2009) New Revision: 1114 Modified: cpu/x86/pc/olpc/mfgdata.fth Log: OLPC - improved the pagination for manufacturing data display in the presence of long key override tags. Modified: cpu/x86/pc/olpc/mfgdata.fth =================================================================== --- cpu/x86/pc/olpc/mfgdata.fth 2009-02-20 21:43:04 UTC (rev 1113) +++ cpu/x86/pc/olpc/mfgdata.fth 2009-02-20 21:44:09 UTC (rev 1114) @@ -90,6 +90,7 @@ dup >r blue-letters bounds ?do + #out @ 3 + rmargin @ > if exit? ?leave then 3 .tab i c@ <# u# u# u#> type loop black-letters

11 years, 5 months

r1113 - cpu/x86/pc/olpc

by svn＠openfirmware.info

Author: wmb Date: 2009-02-20 22:43:04 +0100 (Fri, 20 Feb 2009) New Revision: 1113 Modified: cpu/x86/pc/olpc/smbios.fth Log: OLPC SMBIOS - cleaned up a few things that were misreported. Modified: cpu/x86/pc/olpc/smbios.fth =================================================================== --- cpu/x86/pc/olpc/smbios.fth 2009-02-20 19:01:27 UTC (rev 1112) +++ cpu/x86/pc/olpc/smbios.fth 2009-02-20 21:43:04 UTC (rev 1113) @@ -77,7 +77,7 @@ ( 06 ) h# f000 w, \ BIOS starting address segment ( 08 ) 3 c, \ Release date index ( 09 ) h# 0f c, \ BIOS ROM size in 64K chunks, minus 1 -( 0a ) h# 91880 l, \ BIOS characteristics - PCI, Reflash, selectable boot, EDD +( 0a ) h# 99880 l, \ BIOS characteristics - PCI, Reflash, shadowing, CD boot, selectable boot, EDD ( 0e ) 0 l, \ Vendor and system specific BIOS characteristics ( 12 ) 1 c, \ ACPI is supported \ " OLPC" c$, \ Vendor string @@ -92,12 +92,13 @@ ( 04 ) 1 c, \ Manufacturer string index ( 05 ) 2 c, \ Product name string index ( 06 ) 3 c, \ Version string index -( 07 ) 0 c, \ Serial number string index +( 07 ) 4 c, \ Serial number string index ( 08 ) here to uuid-adr h# 10 allot \ SETME ( 18 ) 6 c, \ Wake up type \ " OLPC" c$, \ 1: Manufacturer \ " XO" c$, \ 2: Product Name \ " 1" c$, \ 3: Version +\ " <sn>" c$, \ 4: Serial number \ 0 c, create base-board-info @@ -108,7 +109,7 @@ ( 05 ) 2 c, \ Product string index ( 06 ) 3 c, \ Version string index ( 07 ) 4 c, \ Serial number string index -\ " OLPC" c$, \ 1: Manufacturer +\ " QUANTA" c$, \ 1: Manufacturer \ " XO" c$, \ 2: Product Name \ " 1" c$, \ 3: Version @@ -157,7 +158,7 @@ ( 01 ) h# 13 c, \ Length ( 02 ) h# 0701 w, \ Handle ( 04 ) 0 c, \ Socket string index -( 05 ) h# 181 w, \ Writeback, enabled, internal, not socketed, L1 +( 05 ) h# 180 w, \ Writeback, enabled, internal, not socketed, L1 ( 07 ) h# 8002 w, \ Max size - 128K ( 09 ) h# 8002 w, \ Installed size - 128K ( 0b ) 1 w, \ Supported SRAM type - unknown @@ -174,7 +175,7 @@ ( 01 ) h# 13 c, \ Length ( 02 ) h# 0702 w, \ Handle ( 04 ) 0 c, \ Socket string index -( 05 ) h# 181 w, \ Writeback, enabled, internal, not socketed, L1 +( 05 ) h# 180 w, \ Writeback, enabled, internal, not socketed, L1 ( 07 ) h# 8001 w, \ Max size - 64K ( 09 ) h# 8001 w, \ Installed size - 64K ( 0b ) 1 w, \ Supported SRAM type - unknown @@ -191,7 +192,7 @@ ( 01 ) h# 13 c, \ Length ( 02 ) h# 0703 w, \ Handle ( 04 ) 0 c, \ Socket string index -( 05 ) h# 182 w, \ Writeback, enabled, internal, not socketed, L2 +( 05 ) h# 181 w, \ Writeback, enabled, internal, not socketed, L2 ( 07 ) h# 8002 w, \ Max size - 128K ( 09 ) h# 8002 w, \ Installed size - 128K ( 0b ) 1 w, \ Supported SRAM type - unknown @@ -396,10 +397,11 @@ +OLPC " XO" +smbios$ " 1" +smbios$ \ Version + " SN" get-tag$ +smbios$ end-smbios-table base-board-info copy-smbios-table ( adr ) - +OLPC + " QUANTA" +smbios$ " XO" +smbios$ " 1" +smbios$ \ Version " SN" get-tag$ +smbios$

11 years, 5 months

r1112 - cpu/x86/pc/olpc

by svn＠openfirmware.info

Author: wmb Date: 2009-02-20 20:01:27 +0100 (Fri, 20 Feb 2009) New Revision: 1112 Modified: cpu/x86/pc/olpc/fw.bth Log: OLPC - turned off "stand-init-debug" to reduce the verbosity of the startup messages on the serial port. The messages that it eliminates are not associated with any common failure modes, so that particular verbosity is pointless. Modified: cpu/x86/pc/olpc/fw.bth =================================================================== --- cpu/x86/pc/olpc/fw.bth 2009-02-20 18:59:25 UTC (rev 1111) +++ cpu/x86/pc/olpc/fw.bth 2009-02-20 19:01:27 UTC (rev 1112) @@ -233,8 +233,8 @@ ; fload ${BP}/dev/geode/usb.fth -\ false to stand-init-debug? -true to stand-init-debug? +false to stand-init-debug? +\ true to stand-init-debug? hex : i-key-wait ( ms -- pressed? )

11 years, 5 months

r1111 - dev/olpc/spiflash

by svn＠openfirmware.info

Author: wmb Date: 2009-02-20 19:59:25 +0100 (Fri, 20 Feb 2009) New Revision: 1111 Modified: dev/olpc/spiflash/spiui.fth Log: OLPC reflasher - use verify-firmware instead of verify after reflashing. It makes no difference for the normal OLPC build, but it prevents spurious verify failures in the custom build case where you enable config variable storage in SPI FLASH. Modified: dev/olpc/spiflash/spiui.fth =================================================================== --- dev/olpc/spiflash/spiui.fth 2009-02-20 18:56:33 UTC (rev 1110) +++ dev/olpc/spiflash/spiui.fth 2009-02-20 18:59:25 UTC (rev 1111) @@ -265,11 +265,11 @@ write-firmware spi-us d# 20 < if - ['] verify catch if + ['] verify-firmware catch if ." Verify failed. Retrying once" cr spi-identify write-firmware - verify + verify-firmware then flash-write-disable else

11 years, 5 months

r1110 - cpu/x86/pc

by svn＠openfirmware.info

Author: wmb Date: 2009-02-20 19:56:33 +0100 (Fri, 20 Feb 2009) New Revision: 1110 Modified: cpu/x86/pc/biosints.fth Log: Biosints - turned off the "unimplemented INT" messages for the two functions that we intentionally don't implement. Modified: cpu/x86/pc/biosints.fth =================================================================== --- cpu/x86/pc/biosints.fth 2009-02-20 18:16:26 UTC (rev 1109) +++ cpu/x86/pc/biosints.fth 2009-02-20 18:56:33 UTC (rev 1110) @@ -839,8 +839,9 @@ h# c1 of 'ebda 4 rshift rm-es! endof \ Segment address of extended BIOS data area h# c2 of handle-mouse endof h# e8 of bigmem-int endof -\ h# e9 of endof \ Don't know what this is. Ralf Brown's interrupt list says -\ \ PhysTechSoft PTS ROM-DOS, but I doubt that is right + h# e9 of rm-set-cf endof \ Don't know what this is. Ralf Brown's interrupt list says + \ PhysTechSoft PTS ROM-DOS, but I doubt that is right + \ Windows invokes it but seems to be okay with it failing ( default ) rm-set-cf ." Unsupported INT 15 AH=" dup . cr endcase @@ -927,7 +928,10 @@ h# 0b of rm-cl@ cfgadr config-b! endof h# 0c of rm-cx@ cfgadr config-w! endof h# 0d of rm-ecx@ cfgadr config-l! endof + + h# 0e of h# 81 rm-ah! rm-set-cf endof \ Fail PCI interrupt routing for now \ h# 0e of pci-int-rout endof + \ h# 0f of set-pci-int endof ( default ) h# 81 rm-ah! rm-set-cf

11 years, 5 months

r1109 - cpu/x86/pc/olpc

by svn＠openfirmware.info

Author: wmb Date: 2009-02-20 19:16:26 +0100 (Fri, 20 Feb 2009) New Revision: 1109 Modified: cpu/x86/pc/olpc/smbios.fth Log: SMBIOS - fixed the UUID routine. Modified: cpu/x86/pc/olpc/smbios.fth =================================================================== --- cpu/x86/pc/olpc/smbios.fth 2009-02-20 09:17:56 UTC (rev 1108) +++ cpu/x86/pc/olpc/smbios.fth 2009-02-20 18:16:26 UTC (rev 1109) @@ -337,23 +337,26 @@ ; : get-tag$ ( tag$ -- value$ ) find-tag 0= if " Not Available" then ?-null ; -: too-long? ( dst-adr -- dst-adr flag ) dup pad - d# 10 >= ; +: too-long? ( dst-adr -- dst-adr flag ) dup pad - h# 10 >= ; : (uuid) ( -- true | adr len false ) " U#" find-tag if ( adr len ) + ?-null ( adr len' ) pad -rot ( dst-adr adr len ) bounds ?do ( dst-adr ) too-long? if drop true unloop exit then i c@ h# 10 digit if ( dst-adr digith ) i 1+ c@ h# 10 digit if ( dst-adr digith digitl ) - swap bwjoin over c! ( dst-adr ) - 1+ ( dst-adr' ) + swap 4 lshift or ( dst-adr byte ) + over c! 1+ ( dst-adr' ) + 2 ( dst-adr advance ) else ( dst-adr digith char ) 3drop true unloop exit then ( dst-adr ) else ( dst-adr char ) [char] - <> if drop true unloop exit then - then ( dst-adr ) - loop ( dst-adr ) + 1 ( dst-adr advance ) + then ( dst-adr advance ) + +loop ( dst-adr ) pad tuck - ( adr len ) dup h# 10 = if ( adr len ) false \ Good UUID ( adr len false )

11 years, 5 months

r1108 - in cpu/x86/pc: biosload olpc

by svn＠openfirmware.info

Author: wmb Date: 2009-02-20 10:17:56 +0100 (Fri, 20 Feb 2009) New Revision: 1108 Modified: cpu/x86/pc/biosload/c32vesa.fth cpu/x86/pc/biosload/callvbe.fth cpu/x86/pc/biosload/fw.bth cpu/x86/pc/olpc/iflash.fth Log: Biosload version - improvements to VESA BIOS handling under Syslinux. Modified: cpu/x86/pc/biosload/c32vesa.fth =================================================================== --- cpu/x86/pc/biosload/c32vesa.fth 2009-02-20 09:14:16 UTC (rev 1107) +++ cpu/x86/pc/biosload/c32vesa.fth 2009-02-20 09:17:56 UTC (rev 1108) @@ -1,8 +1,8 @@ purpose: Call VESA BIOS functions via syslinux c32 callback -[ifdef] notdef-get-com32-ptr -\ sp 0 #) mov \ Save COM32 stack pointer to get parameters -[then] +\ [ifdef] notdef-get-com32-ptr + sp 0 #) mov \ Save COM32 stack pointer for possible later use +\ [then] [ifdef] notdef-getvbe \ There is an extra return address on the stack from the e9 call above d# 16 [sp] bp mov \ COM32 intcall helper function Modified: cpu/x86/pc/biosload/callvbe.fth =================================================================== --- cpu/x86/pc/biosload/callvbe.fth 2009-02-20 09:14:16 UTC (rev 1107) +++ cpu/x86/pc/biosload/callvbe.fth 2009-02-20 09:17:56 UTC (rev 1108) @@ -9,12 +9,12 @@ d# 16 [ax] bx mov \ COM32 intcall helper function d# 20 [ax] dx mov \ bounce buffer address - 4f02 # d# 36 [dx] mov \ AX + h# 4f02 # d# 36 [dx] mov \ AX cx d# 24 [dx] mov \ BX - \ dx push - 0 # push dx push +\ 0 # push + dx push h# 10 # push bx call @@ -22,3 +22,118 @@ bp pop di pop si pop c; + +: +c32-regs ( offset -- adr ) 0 @ d# 20 + @ + ; +: 'c32-gs ( -- adr ) 0 +c32-regs ; +: 'c32-fs ( -- adr ) 2 +c32-regs ; +: 'c32-es ( -- adr ) 4 +c32-regs ; +: 'c32-ds ( -- adr ) 6 +c32-regs ; + +: 'c32-di ( -- adr ) d# 8 +c32-regs ; +: 'c32-si ( -- adr ) d# 12 +c32-regs ; +: 'c32-bp ( -- adr ) d# 16 +c32-regs ; +: 'c32-bx ( -- adr ) d# 24 +c32-regs ; +: 'c32-dx ( -- adr ) d# 28 +c32-regs ; +: 'c32-cx ( -- adr ) d# 32 +c32-regs ; +: 'c32-ax ( -- adr ) d# 36 +c32-regs ; +: 'c32-eflags ( -- adr ) d# 40 +c32-regs ; + +code c32-intcall ( int# -- ) + cx pop + + si push di push bp push + 0 #) ax mov \ Pointer to COM32 args + d# 16 [ax] bx mov \ COM32 intcall helper function + d# 20 [ax] dx mov \ bounce buffer address + + dx push +\ 0 # push + dx push + cx push \ Int# + + bx call + ax pop ax pop ax pop + + bp pop di pop si pop +c; + +: vbe-call ( function# -- ) + 'c32-ax l! h# 10 c32-intcall + 'c32-ax l@ h# 4f <> abort" VESA BIOS call failed" +; + +: c32-es:di! ( adr -- ) >seg:off 'c32-es w! 'c32-di w! ; +: vbe-info ( -- adr ) + h# 200 +c32-regs ( adr ) + h# 32454256 over l! ( adr ) + c32-es:di! + h# 4f00 vbe-call + h# 200 +c32-regs ( adr ) +; +: .vesa-modes ( -- ) + vbe-info ( adr ) + dup l@ h# 41534556 <> if ( adr ) + ." VBE info call failed" cr + drop exit + then ( adr ) + dup 6 + seg:off@ .cstr cr ( adr ) + d# 14 + seg:off@ ( 'mode-list ) + begin dup w@ dup h# ffff <> while . wa1+ repeat cr 2drop +; +: vesa-mode-info ( mode# -- adr ) + 'c32-cx l! ( ) + h# 200 +c32-regs c32-es:di! ( ) + h# 4f01 vbe-call ( ) + h# 200 +c32-regs ( adr ) +; + +[ifdef] Commentary +00.w mode attributes +02.b wina attributes +03.b winb attributes +04.w winGranularity +06.w WinSize +08.w WinASegment +0a.w WinBSegment +0c.l WinFuncPtr +10.w BytesPerScanLine +12.w XResolution +14.w YResolution +16.b XCharSize +17.b YCharSize +18.b NumberOfPlanes +19.b BitsPerPixel +1a.b NumberOfBanks +1b.b MemoryModel +1c.b BankSize +1d.b NumberOfImagePlanes +1e.b Reserved +1f.b RedMaskSize +20.b RedFieldPosition +21.b GreenMaskSize +22.b GreenFieldPosition +23.b BlueMaskSize +24.b BlueFieldPosition +25.b RsvdMaskSize +26.b RsvdFieldPosition +27.b DirectColorModeInfo +\ Following are VBE 2.0 and above +28.l PhysBasePtr +2c.l Reserved +30.w Reserved +\ Following are VBE 3.0 and above +32.w LinBytesPerScanLine +34.b BnkNumberOfImagePages +35.b LinNumberOfImagePages +36.b LinRedMaskSize +37.b LinRedFieldPosition +38.b LinGreenMaskSize +39.b LinGreenFieldPosition +3a.b LinBlueMaskSize +3b.b LinBlueFieldPosition +3c.b LinRsvdMaskSize +3d.b LinRsvdFieldPosition +3e.l MaxPixelClock +[then] + +: current-vesa-mode ( -- mode# ) h# 4f03 vbe-call 'c32-bx l@ ; Modified: cpu/x86/pc/biosload/fw.bth =================================================================== --- cpu/x86/pc/biosload/fw.bth 2009-02-20 09:14:16 UTC (rev 1107) +++ cpu/x86/pc/biosload/fw.bth 2009-02-20 09:17:56 UTC (rev 1108) @@ -226,6 +226,10 @@ fload ${BP}/cpu/x86/pc/biosload/callbios.fth \ fload ${BP}/cpu/x86/pc/biosload/rmenter.fth +[ifdef] syslinux-loaded +fload ${BP}/cpu/x86/pc/biosload/callvbe.fth +[then] + \ false to stand-init-debug? true to stand-init-debug? Modified: cpu/x86/pc/olpc/iflash.fth =================================================================== --- cpu/x86/pc/olpc/iflash.fth 2009-02-20 09:14:16 UTC (rev 1107) +++ cpu/x86/pc/olpc/iflash.fth 2009-02-20 09:17:56 UTC (rev 1108) @@ -22,6 +22,7 @@ /mfg-data-merge move ; +[ifdef] get-mfg-data \ unpollute-mfg-data clears the Insyde BIOS residue from the beginning \ of the manufacturing data area, restoring it to erased state : unpollute-mfg-data ( -- ) @@ -30,6 +31,7 @@ mfg-data-buf mfg-data-merge-offset h# ff fill put-mfg-data ; +[then] : get-insyde ( ["filename"] -- ) parse-word ( adr len )

11 years, 5 months

r1107 - cpu/x86/pc cpu/x86/pc/olpc dev/geode/display

by svn＠openfirmware.info

Author: wmb Date: 2009-02-20 10:14:16 +0100 (Fri, 20 Feb 2009) New Revision: 1107 Modified: cpu/x86/pc/biosints.fth cpu/x86/pc/olpc/dsdt.dsl dev/geode/display/gxvga.fth Log: OLPC - Fixed the VESA BIOS emulation for Windows, supporting mode 112 (640x480x32), 115 (800x600x32), and 118 (1024x768x32) Modified: cpu/x86/pc/biosints.fth =================================================================== --- cpu/x86/pc/biosints.fth 2009-02-20 08:23:30 UTC (rev 1106) +++ cpu/x86/pc/biosints.fth 2009-02-20 09:14:16 UTC (rev 1107) @@ -200,18 +200,52 @@ create mode12-info \ w b b w w w w d w w w b b b b b b b b b h# db w, \ Linear, (VGA), Graphics, Color, no TTY Output, D1=1, hardware-supported +create mode-112-info +\ h# bf w, \ Linear, VGA, Graphics, Color, TTY Output, D1=1, hardware-supported +\ 7 c, 0 c, d# 64 w, d# 64 w, h# a000 w, 0 w, 0 l, \ Not windowed + h# fb w, \ Linear, noVGA, Graphics, Color, noTTY Output, D1=1, hardware-supported + 0 c, 0 c, d# 0 w, d# 0 w, 0 w, 0 w, 0 l, \ Not windowed + d# 640 /l* w, \ BytesPerScanLine + d# 640 w, d# 480 w, \ X, Y res + d# 8 c, d# 16 c, \ Char width, height + 1 c, \ NumPlanes + d# 32 c, \ Bits/pixel + 1 c, \ NumBanks + 6 c, \ MemModel - DirectColor + 0 c, \ Bank size (not banked) + 2 c, \ NumImagePages + 1 c, \ Reserved + + \ Banked Color info + 8 c, d# 16 c, 8 c, 8 c, 8 c, 0 c, 0 c, 0 c, \ {RGBX}{Bits,Pos} + 0 c, \ Gamma fixed (change if we implement function 9) + + fb-pci-base l, 0 l, 0 w, \ Framebuffer address + + \ Linear info + d# 640 /l* w, \ Bytes per scan line + 0 c, \ No banks + +\ fbsize d# 640 / d# 480 / 4 / c, ( need to account for cmd buffer ) + d# 11 c, \ Number of images that will fit in framebuffer + + 8 c, d# 16 c, 8 c, 8 c, 8 c, 0 c, 0 c, 0 c, \ {RGBX}{Bits,Pos} + + d# 56,200,000 l, \ Max pixel clock +here mode-112-info - constant /mode-112-info + create mode-115-info - h# fb w, \ Linear, NotVGA, Graphics, Color, no TTY Output, D1=1, hardware-supported - 0 c, 0 c, 0 w, d# 64 w, 0 w, 0 w, 0 l, \ Not windowed - 0 w, \ BytesPerScanLine irrelevant in linear mode + h# fb w, \ Linear, VGA, Graphics, Color, TTY Output, D1=1, hardware-supported + 0 c, 0 c, d# 0 w, d# 0 w, h# 0 w, 0 w, 0 l, \ Not windowed + d# 800 /l* w, \ BytesPerScanLine d# 800 w, d# 600 w, \ X, Y res - d# 15 c, d# 18 w, \ Char width, height (sort of irrelevant) - 1 c, \ NumPlanes (irrelevant for linear?) - d# 24 c, \ Bits/pixel + d# 8 c, d# 16 c, \ Char width, height + 1 c, \ NumPlanes + d# 32 c, \ Bits/pixel 1 c, \ NumBanks - 6 c, \ MemModel - DirectColor (?) + 6 c, \ MemModel - DirectColor 0 c, \ Bank size (not banked) - 0 c, \ NumImagePages + 2 c, \ NumImagePages 1 c, \ Reserved \ Banked Color info @@ -221,11 +255,11 @@ fb-pci-base l, 0 l, 0 w, \ Framebuffer address \ Linear info - d# 4096 /w* w, \ Bytes per scan line + d# 800 /l* w, \ Bytes per scan line 0 c, \ No banks -\ fbsize d# 1200 / d# 900 / 2/ c, - 3 c, \ Number of images that will fit in framebuffer +\ fbsize d# 800 / d# 600 / 4/ c, + 7 c, \ Number of images that will fit in framebuffer 8 c, d# 16 c, 8 c, 8 c, 8 c, 0 c, 0 c, 0 c, \ {RGBX}{Bits,Pos} @@ -233,17 +267,18 @@ here mode-115-info - constant /mode-115-info create mode-118-info - h# fb w, \ Linear, NotVGA, Graphics, Color, no TTY Output, D1=1, hardware-supported - 0 c, 0 c, 0 w, d# 64 w, 0 w, 0 w, 0 l, \ Not windowed - 0 w, \ BytesPerScanLine irrelevant in linear mode + h# fb w, \ Linear, VGA, Graphics, Color, TTY Output, D1=1, hardware-supported + 0 c, 0 c, d# 0 w, d# 0 w, h# 0 w, 0 w, 0 l, \ Not windowed + d# 1024 /l* w, \ BytesPerScanLine d# 1024 w, d# 768 w, \ X, Y res - d# 15 c, d# 18 w, \ Char width, height (sort of irrelevant) - 1 c, \ NumPlanes (irrelevant for linear?) - d# 24 c, \ Bits/pixel +\ d# 12 c, d# 15 c, \ Char width, height + d# 8 c, d# 16 c, \ Char width, height (sort of irrelevant) + 1 c, \ NumPlanes + d# 32 c, \ Bits/pixel 1 c, \ NumBanks - 6 c, \ MemModel - DirectColor (?) + 6 c, \ MemModel - DirectColor 0 c, \ Bank size (not banked) - 0 c, \ NumImagePages + 2 c, \ NumImagePages 1 c, \ Reserved \ Banked Color info @@ -253,11 +288,11 @@ fb-pci-base l, 0 l, 0 w, \ Framebuffer address \ Linear info - d# 4096 /w* w, \ Bytes per scan line + d# 1024 /l* w, \ Bytes per scan line 0 c, \ No banks -\ fbsize d# 1200 / d# 900 / 2/ c, - 3 c, \ Number of images that will fit in framebuffer +\ fbsize d# 1024 / d# 768 / 4 / c, + 4 c, \ Number of images that will fit in framebuffer 8 c, d# 16 c, 8 c, 8 c, 8 c, 0 c, 0 c, 0 c, \ {RGBX}{Bits,Pos} @@ -266,11 +301,11 @@ create mode-120-info h# fb w, \ Linear, NotVGA, Graphics, Color, no TTY Output, D1=1, hardware-supported - 0 c, 0 c, 0 w, d# 64 w, 0 w, 0 w, 0 l, \ Not windowed - 0 w, \ BytesPerScanLine irrelevant in linear mode + 0 c, 0 c, d# 0 w, d# 0 w, h# 0 w, 0 w, 0 l, \ Not windowed + d# 1200 /w* w, \ BytesPerScanLine d# 1200 w, d# 900 w, \ X, Y res - d# 15 c, d# 18 w, \ Char width, height (sort of irrelevant) - 1 c, \ NumPlanes (irrelevant for linear?) + d# 15 c, d# 18 c, \ Char width, height (sort of irrelevant) + 1 c, \ NumPlanes d# 16 c, \ Bits/pixel 1 c, \ NumBanks 6 c, \ MemModel - DirectColor (?) @@ -289,7 +324,7 @@ 0 c, \ No banks \ fbsize d# 1200 / d# 900 / 2/ c, - 7 c, \ Number of images that will fit in framebuffer + 6 c, \ Number of images that will fit in framebuffer 5 c, d# 11 c, 6 c, 5 c, 5 c, 0 c, 0 c, 0 c, \ {RGBX}{Bits,Pos} @@ -309,9 +344,11 @@ : vbe-ok ( -- ) h# 4f rm-ax! ; : vbe-modes ( -- ) +\ ." vbe-modes" cr ?vbe2 h# 41534556 0 >vbe-pa l! \ VbeSignature - h# 0300 4 >vbe-pa w! \ VbeVersion +\ h# 0300 4 >vbe-pa w! \ VbeVersion + h# 0200 4 >vbe-pa w! \ VbeVersion 1 d# 10 >vbe-pa l! \ Capabilities - 8-bit DACs vbebuf d# 34 + d# 14 >vbe-pa seg:off! \ VbeFarPtr to mode list fbsize d# 16 rshift d# 18 >vbe-pa w! \ TotalMemory @@ -328,33 +365,56 @@ d# 34 >vbe-pa \ 3 w!++ \ Text mode 3 \ h# 12 w!++ \ Graphics mode 12 - h# 115 w!++ \ 800x600x24 - h# 118 w!++ \ 1024x768x24 - h# 120 w!++ \ OLPC native mode + h# 112 w!++ \ 640x480x32 + h# 115 w!++ \ 800x600x32 + h# 118 w!++ \ 1024x768x32 +\ h# 120 w!++ \ OLPC native mode -1 swap w! \ End of list vbe-ok ; : vbe-get-mode ( -- ) - rm-cx@ case - h# 115 of mode-115-info /mode-115-info endof - h# 118 of mode-118-info /mode-118-info endof - h# 120 of mode-120-info /mode-120-info endof + + rm-cx@ +\ ." vbe-get-mode " dup . cr +h# 1ff and case + h# 112 of mode-112-info /mode-112-info endof + h# 115 of mode-115-info /mode-115-info endof + h# 118 of mode-118-info /mode-118-info endof +\ h# 120 of mode-120-info /mode-120-info endof ( default ) ." Bad VBE mode number " dup . cr 0 0 rot endcase ( adr len ) 0 >vbe-pa h# 100 erase 0 >vbe-pa swap move + vbe-ok ; +: vbe-error ( -- ) h# 014f rm-ax! ; +0 value vbe-this-mode : vbe-set-mode ( -- ) - ." VBE set mode " rm-cx@ . cr - debug-me + rm-bx@ case + h# 4112 of " 640x480x32" endof + h# 4115 of " 800x600x32" endof + h# 4118 of " 1024x768x32" endof + ( default ) drop vbe-error exit + endcase + + " screen-ih" eval ['] $call-method catch if + 3drop vbe-error exit + then + rm-bx@ to vbe-this-mode + vbe-ok ; +: vbe-current-mode ( -- ) + vbe-this-mode rm-bx! + vbe-ok +; : vesa-bios ( -- ) rm-al@ case h# 00 of vbe-modes endof h# 01 of vbe-get-mode endof h# 02 of vbe-set-mode endof + h# 03 of vbe-current-mode endof ( default ) ." Unsupported VBE function" dup .x cr endcase ; Modified: cpu/x86/pc/olpc/dsdt.dsl =================================================================== --- cpu/x86/pc/olpc/dsdt.dsl 2009-02-20 08:23:30 UTC (rev 1106) +++ cpu/x86/pc/olpc/dsdt.dsl 2009-02-20 09:14:16 UTC (rev 1107) @@ -771,7 +771,6 @@ Name (_UID, One) Method (_CRS, 0, NotSerialized) { Name (MBRB, ResourceTemplate () { - Memory32Fixed (ReadWrite, 0x00000000, 0x000A0000, ) Memory32Fixed (ReadOnly, 0x000E0000, 0x00020000, ) Memory32Fixed (ReadWrite, 0x00100000, 0x00000000, _Y06) // Edited below // Assumes that the SMM memory is at 8040.0000, I think @@ -781,7 +780,6 @@ IO (Decode16, 0x0092, 0x0092, 0x00, 0x01, ) IO (Decode16, 0x0030, 0x0030, 0x00, 0x10, ) // I/O ports to fake out SMI interrupts - IO (Decode16, 0x03c0, 0x03c0, 0x00, 0x20, ) // Claim VGA I/O ports // We claim the EC ports here instead of in an EC node // because we don't want Windows to load an EC driver // nor to complain about not being able to do so. @@ -839,7 +837,7 @@ Name (_HID, EisaId ("PNP0501")) Name (_UID, One) Name (_DDN, "COM1") - Name (_STA, 3) // Present and decoding resources, but shouldn't be shown + Name (_STA, 0xf) // You can use the serial port if you open the box Name (_CRS, ResourceTemplate () { IO (Decode16, 0x03F8, 0x03F8, 0x00, 0x08, ) IRQNoFlags () {4} Modified: dev/geode/display/gxvga.fth =================================================================== --- dev/geode/display/gxvga.fth 2009-02-20 08:23:30 UTC (rev 1106) +++ dev/geode/display/gxvga.fth 2009-02-20 09:14:16 UTC (rev 1107) @@ -554,26 +554,38 @@ 1 ms \ Let memory accesses finish 4 dc@ 1 invert and 4 dc! \ Turn off FIFO load ; -: 8bpp ( -- ) 8 dc@ h# 300 invert and 8 dc! ; \ drop down to 8bpp mode +: 8bpp ( -- ) 8 dc@ h# 300 invert and 8 dc! ; +: 16bpp ( -- ) 8 dc@ h# 300 invert and h# 100 or 8 dc! ; +: 24bpp ( -- ) 8 dc@ h# 300 invert and h# 200 or 8 dc! ; +: 32bpp ( -- ) 8 dc@ h# 300 or 8 dc! ; : vga-start ( -- ) 8 dc@ 1 or 8 dc! \ Timing generator \ Fixed timing must be off to use the upscaler 4 dc@ h# 40000 invert and h# 80 or 4 dc! \ Fixed Timing (40000) off, VGA (80) on ; -: display-lfb ( -- ) 5601 4 dc! ; \ Enable linear framebuffer +: display-lfb ( -- ) 6501 4 dc! ; \ Enable linear framebuffer -: setup-filter-mode3 ( -- ) -\ d# 400 1- d# 640 1- wljoin h# 5c dc! +: setup-filter-720x400 ( -- ) d# 400 1- d# 720 1- wljoin h# 5c dc! h# 1c802666 h# 90 dc! filter-taps ; -: setup-filter-mode12 ( -- ) +: setup-filter-640x480 ( -- ) d# 480 1- d# 640 1- wljoin h# 5c dc! h# 22222222 h# 90 dc! filter-taps ; +: setup-filter-800x600 ( -- ) + d# 600 1- d# 800 1- wljoin h# 5c dc! + h# 2ab42ab4 h# 90 dc! + filter-taps +; +: setup-filter-1024x768 ( -- ) + d# 768 1- d# 1024 1- wljoin h# 5c dc! + h# 369d369d h# 90 dc! + filter-taps +; : filteron ( -- ) h# 1000 h# 94 dc! ; : filteroff ( -- ) h# 40004000 h# 90 dc! @@ -629,7 +641,7 @@ \ otherwise the filter and the VGA don't synchronize to the \ frame and the sequencer shuts off after one frame. \ That can be fixed with all-off vga-start - setup-filter-mode3 filteron + setup-filter-720x400 filteron all-off 8bpp vga-start use-vga-dac @@ -642,7 +654,7 @@ \ otherwise the filter and the VGA don't synchronize to the \ frame and the sequencer shuts off after one frame. \ That can be fixed with all-off vga-start - setup-filter-mode12 filteron + setup-filter-640x480 filteron all-off 8bpp vga-start use-vga-dac @@ -657,6 +669,36 @@ : graphics-mode12 ( -- ) switch-to-vga-mode12 graphics-mode12 ; +: 640x480x32 ( -- ) \ vesa-mode112 + set-mode \ Get back into native resolution linear mode + unlock + 32bpp + d# 640 d# 32 * 6 rshift dup h# 30 dc! h# 34 dc! + all-off setup-filter-640x480 filteron + display-on + lock +; + +: 800x600x32 ( -- ) \ vesa-mode115 + set-mode \ Get back into native resolution linear mode + unlock + 32bpp + d# 800 d# 32 * 6 rshift dup h# 30 dc! h# 34 dc! + all-off setup-filter-800x600 filteron + display-on + lock +; + +: 1024x768x32 ( -- ) \ vesa-mode118 + set-mode \ Get back into native resolution linear mode + unlock + 32bpp + d# 1024 d# 32 * 6 rshift dup h# 30 dc! h# 34 dc! + all-off setup-filter-1024x768 filteron + display-on + lock +; + warning ! \ LICENSE_BEGIN

11 years, 5 months

r1106 - cpu/x86/pc/olpc

by svn＠openfirmware.info

Author: wmb Date: 2009-02-20 09:23:30 +0100 (Fri, 20 Feb 2009) New Revision: 1106 Modified: cpu/x86/pc/olpc/vsapci.fth Log: OLPC - Fixed the longstanding problem whereby the Windows XO Display Driver wouldn't claim the legacy resources. The problem was that the "IO Space Enable" bit in the virtual config register for the display was off. That made the Windows VideoPort driver refuse to claim the resources. Modified: cpu/x86/pc/olpc/vsapci.fth =================================================================== --- cpu/x86/pc/olpc/vsapci.fth 2009-02-12 08:26:53 UTC (rev 1105) +++ cpu/x86/pc/olpc/vsapci.fth 2009-02-20 08:23:30 UTC (rev 1106) @@ -157,7 +157,7 @@ ff800000 , ffffc000 , ffffc000 , ffffc000 , 0 , 0 , 0 , 0 , - 30100b , 2200002 , 3000000 , 8 , + 30100b , 2200003 , 3000000 , 8 , fb-pci-base , gp-pci-base , dc-pci-base , vp-pci-base , \ FB, GP, DC, VP 0 , 0 , 0 , 30100b , \ VIP (LX only) 0 , 0 , 0 , 0 ,

11 years, 5 months

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

openfirmware February 2009