September 2011 - openfirmware - mail.coreboot.org

[commit] r2525 - ofw/inet
by repository service Sept. 22, 2011

Sept. 22, 2011

Author: wmb Date: Fri Sep 23 00:38:28 2011 New Revision: 2525 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/2525 Log: Fixed pingd, which has been broken for a long time. Modified: ofw/inet/ping.fth Modified: ofw/inet/ping.fth ============================================================================== --- ofw/inet/ping.fth Wed Sep 21 19:16:42 2011 (r2524) +++ ofw/inet/ping.fth Fri Sep 23 00:38:28 2011 (r2525) @@ -30,23 +30,17 @@ /packet-max buffer: packet : get-packet? ( -- packet? ) - packet /packet-max " read" $call-net to /packet - /packet -1 = if cr ." Packet error" cr false exit then - /packet -2 = if false exit then - ." ." true + h# 800 " receive-ethernet-packet" $call-net if false exit then ( adr len ) + /packet-max min to /packet ( adr ) + packet /packet move ( ) + ." ." true ( packet ) ; d# 14 constant /ether-header 0 value ip-offset : ip-header ( -- adr ) packet ip-offset + ; -: ip? ( -- flag ) - ip-header c@ h# 45 = if - ip-header - then - - false -; +: ip? ( -- flag ) ip-header c@ h# 45 = ; : link-level-ok? ( -- flag ) packet c@ h# 45 = if @@ -110,7 +104,10 @@ packet packet 6 + 6 exchange-bytes then ; -: exchange-ips ( -- ) ip-header d# 12 + dup 4 + 4 exchange-bytes ; +: 'ip-src ( -- adr ) ip-header d# 12 + ; +: 'ip-dst ( -- adr ) ip-header d# 16 + ; + +: exchange-ips ( -- ) 'ip-src 'ip-dst 4 exchange-bytes ; : change-type ( -- ) 0 ip-payload drop c! ; : recompute-ip-checksum ( -- ) 0 ip-header d# 10 + be-w! \ Zap IP checksum @@ -148,8 +145,7 @@ ; : send-packet ( -- ) - packet /packet " write" $call-net - /packet <> if cr ." Send failed" cr then + packet /packet 'ip-dst h# 800 " send-link-packet" $call-net ; : echo-packet ( -- ) @@ -162,7 +158,6 @@ ; : ?echo-packet ( -- ) ping? if - first? if .ip false to first? then echo-packet then ;

1 0

[commit] r2524 - cpu/arm
by repository service Sept. 21, 2011

Sept. 21, 2011

Author: tooch Date: Wed Sep 21 19:16:42 2011 New Revision: 2524 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/2524 Log: ARM - >resolve broken; should be mask-op-and-or-offset, not simple add. Test: code foo b `0 F:` sub tos,tos,tos 0 L: add tos,tos,tos c; Modified: cpu/arm/assem.fth Modified: cpu/arm/assem.fth ============================================================================== --- cpu/arm/assem.fth Tue Sep 20 08:00:08 2011 (r2523) +++ cpu/arm/assem.fth Wed Sep 21 19:16:42 2011 (r2524) @@ -1438,7 +1438,10 @@ \ These implementation factors are used by the local labels package : <mark ( -- <mark ) here ; : >mark ( -- >mark ) here ; -: >resolve ( >mark -- ) here over >br-offset over asm@ + swap asm! ; +: >resolve ( >mark -- ) + \ Can't use put-branch because this may be a conditional branch. + here over >br-offset over asm@ h# ff000000 land or swap asm! +; : <resolve ( <mark -- <mark ) ; headers

1 0

[commit] r2523 - cpu/arm/olpc/1.75 dev/olpc/spiflash
by repository service Sept. 19, 2011

Sept. 19, 2011

Author: quozl Date: Tue Sep 20 08:00:08 2011 New Revision: 2523 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/2523 Log: OLPC XO-1.75 - hold EC in reset while flashing OpenFirmware SPI FLASH, so that power button does not interrupt write. #11215. Modified: cpu/arm/olpc/1.75/devices.fth cpu/arm/olpc/1.75/fw-version.fth dev/olpc/spiflash/spiui.fth Modified: cpu/arm/olpc/1.75/devices.fth ============================================================================== --- cpu/arm/olpc/1.75/devices.fth Sat Sep 17 00:59:16 2011 (r2522) +++ cpu/arm/olpc/1.75/devices.fth Tue Sep 20 08:00:08 2011 (r2523) @@ -182,16 +182,22 @@ : hdd-led-on ( -- ) d# 10 gpio-set ; : hdd-led-toggle ( -- ) d# 10 gpio-pin@ if hdd-led-off else hdd-led-on then ; -fload ${BP}/dev/olpc/spiflash/spiui.fth \ User interface for SPI FLASH programming -\ fload ${BP}/dev/olpc/spiflash/recover.fth \ XO-to-XO SPI FLASH recovery -: ofw-fw-filename$ " disk:\boot\olpc.rom" ; -' ofw-fw-filename$ to fw-filename$ - fload ${BP}/cpu/arm/olpc/1.75/bbedi.fth fload ${BP}/cpu/arm/olpc/1.75/edi.fth +[ifdef] load-base +: flash-buf load-base ; +[else] +/flash buffer: flash-buf +[then] + fload ${BP}/cpu/arm/olpc/1.75/ecflash.fth +fload ${BP}/dev/olpc/spiflash/spiui.fth \ User interface for SPI FLASH programming +\ fload ${BP}/dev/olpc/spiflash/recover.fth \ XO-to-XO SPI FLASH recovery +: ofw-fw-filename$ " disk:\boot\olpc.rom" ; +' ofw-fw-filename$ to fw-filename$ + 0 0 " d420b000" " /" begin-package " display" name fload ${BP}/cpu/arm/olpc/1.75/lcdcfg.fth Modified: cpu/arm/olpc/1.75/fw-version.fth ============================================================================== --- cpu/arm/olpc/1.75/fw-version.fth Sat Sep 17 00:59:16 2011 (r2522) +++ cpu/arm/olpc/1.75/fw-version.fth Tue Sep 20 08:00:08 2011 (r2523) @@ -1,3 +1,3 @@ \ The overall firmware revision macro: FW_MAJOR B -macro: FW_MINOR 09 +macro: FW_MINOR 09jg Modified: dev/olpc/spiflash/spiui.fth ============================================================================== --- dev/olpc/spiflash/spiui.fth Sat Sep 17 00:59:16 2011 (r2522) +++ dev/olpc/spiflash/spiui.fth Tue Sep 20 08:00:08 2011 (r2523) @@ -42,12 +42,6 @@ \ Perform a series of sanity checks on the new firmware image. -[ifdef] load-base -: flash-buf load-base ; -[else] -/flash buffer: flash-buf -[then] - 0 value file-loaded? : crc ( adr len -- crc ) 0 crctab 2swap ($crc) ; @@ -191,9 +185,23 @@ ." or about 5 minutes if the host is running OFW." cr ; +: ec-spi-reprogrammed ( -- ) + edi-spi-start + set-ec-reboot + unreset-8051 +; + +: ignore-power-button ( -- ) + edi-spi-start + ['] reset-8051 catch if reset-8051 then + use-ssp-spi + ['] ec-spi-reprogrammed to spi-reprogrammed +; + : reflash ( -- ) \ Flash from data already in memory - hdd-led-on ?file + hdd-led-on + ignore-power-button flash-write-enable write-firmware @@ -208,7 +216,7 @@ flash-write-disable else .verify-msg - then + then hdd-led-off ;

1 0

[commit] r2522 - forth/lib
by repository service Sept. 16, 2011

Sept. 16, 2011

Author: wmb Date: Sat Sep 17 00:59:16 2011 New Revision: 2522 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/2522 Log: Forth core - added "dump-audio" command. Modified: forth/lib/dump.fth Modified: forth/lib/dump.fth ============================================================================== --- forth/lib/dump.fth Sat Sep 17 00:58:24 2011 (r2521) +++ forth/lib/dump.fth Sat Sep 17 00:59:16 2011 (r2522) @@ -54,6 +54,21 @@ : dump ( addr len -- ) ['] c@ is dc@ (dump) ; : du ( addr -- addr+64 ) dup d# 64 dump d# 64 + ; +\ Dumps 16-bit signed samples in decimal +: dump-audio ( adr len -- ) + base @ >r + bounds ?do + hex + i 8 u.r ." : " + decimal + i d# 16 bounds do + i <w@ 7 .r + /w +loop + cr + d# 16 +loop + pop-base +; + only forth also definitions \ LICENSE_BEGIN \ Copyright (c) 2006 FirmWorks

1 0

[commit] r2521 - cpu/arm/olpc/1.75 dev/hdaudio
by repository service Sept. 16, 2011

Sept. 16, 2011

Author: wmb Date: Sat Sep 17 00:58:24 2011 New Revision: 2521 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/2521 Log: OLPC XO-1.75 - audio selftest - fixed problem with intermittent channel reversal failures. Modified: cpu/arm/olpc/1.75/sound.fth dev/hdaudio/test.fth Modified: cpu/arm/olpc/1.75/sound.fth ============================================================================== --- cpu/arm/olpc/1.75/sound.fth Fri Sep 16 23:39:35 2011 (r2520) +++ cpu/arm/olpc/1.75/sound.fth Sat Sep 17 00:58:24 2011 (r2521) @@ -156,7 +156,7 @@ 1 h# 80 adma! \ Enable DMA completion interrupts h# 0081.3020 h# 40 adma! \ 16 bits, pack, fetch next, enable, chain, hold dest, inc src ; -: stop-out-ring ( -- ) h# 100000 h# 40 adma! ; +: stop-out-ring ( -- ) h# 100000 h# 40 adma! 0 h# 80 adma! ; : make-in-ring ( -- ) in-desc h# 10 + in-bufs sspa-base /audio-buf in-desc set-descriptor @@ -169,6 +169,7 @@ \ h# 0081.3008 h# 44 adma! \ 16 bits, pack, fetch next, enable, chain, inc dest, hold src h# 00a1.31c8 h# 44 adma! \ 16 bits, pack, fetch next, enable, chain, burst32, inc dest, hold src ; +: stop-in-ring ( -- ) h# 100000 h# 44 adma! 0 h# 84 adma! ; : copy-out ( -- ) my-out-desc >r ( r: desc ) @@ -380,6 +381,7 @@ copy-in ( actual ) repeat ( actual ) disable-sspa-rx ( actual ) + stop-in-ring ( actual ) reset-rx ( actual ) ; : read ( adr len -- actual ) open-in audio-in close-in ; @@ -401,7 +403,9 @@ in-adr0 to in-adr ( ) in-len0 mono? if 2* then to in-len - ( audio-clock-on ) ( ) \ This will mess up any frequency settings + \ Resetting the clock at this point seems to prevent intermittent channel + \ reversal on reception. + audio-clock-on ( ) \ This will mess up any frequency settings setup-sspa-tx ( ) setup-sspa-rx ( ) @@ -428,6 +432,9 @@ disable-sspa-rx ( ) disable-sspa-tx ( ) + stop-in-ring + stop-out-ring + reset-rx reset-tx @@ -523,8 +530,7 @@ false value force-internal-mic? \ Can't be implemented on XO-1.75 2 value #channels -\ Unless you do the audio-clock-on, the L/R phase is often wrong -: input-test-settings ( -- ) ( audio-clock-on ) ; +: input-test-settings ( -- ) ; : output-test-settings ( -- ) ; d# -1 constant case-test-volume Modified: dev/hdaudio/test.fth ============================================================================== --- dev/hdaudio/test.fth Fri Sep 16 23:39:35 2011 (r2520) +++ dev/hdaudio/test.fth Sat Sep 17 00:58:24 2011 (r2521) @@ -183,19 +183,25 @@ instructions-done then ; -\ Returns failure by throwing -: automatic-test ( -- ) - analysis-parameters " set-analysis-parameters" $call-analyzer +: acoustic-test ( -- ) disconnect-loopback \ Not for 1.5; it can test internal while loopback is connected " smt-test?" evaluate if test-with-fixture throw else test-with-case throw then +; +: loopback-test ( -- ) connect-loopback test-with-loopback throw disconnect-loopback ; +\ Returns failure by throwing +: automatic-test ( -- ) + analysis-parameters " set-analysis-parameters" $call-analyzer + acoustic-test + loopback-test +; : selftest ( -- error? ) diagnostic-mode? if open 0= if ." Failed to open /audio" cr true exit then

1 0

[commit] r2520 - cpu/arm/olpc/1.75
by repository service Sept. 16, 2011

Sept. 16, 2011

Author: wmb Date: Fri Sep 16 23:39:35 2011 New Revision: 2520 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/2520 Log: OLPC XO-1.75 - Fixed problem with automatic audio selftest - the channel assignment now works correctly, with no need to flip the sync polarity. Modified: cpu/arm/olpc/1.75/sound.fth Modified: cpu/arm/olpc/1.75/sound.fth ============================================================================== --- cpu/arm/olpc/1.75/sound.fth Fri Sep 16 23:38:23 2011 (r2519) +++ cpu/arm/olpc/1.75/sound.fth Fri Sep 16 23:39:35 2011 (r2520) @@ -405,7 +405,6 @@ setup-sspa-tx ( ) setup-sspa-rx ( ) - active-high-rx-fs ( ) make-in-ring ( ) make-out-ring ( )

1 0

[commit] r2519 - cpu/arm/olpc/1.75
by repository service Sept. 16, 2011

Sept. 16, 2011

Author: wmb Date: Fri Sep 16 23:38:23 2011 New Revision: 2519 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/2519 Log: OLPC XO-1.75 - Goes with last checkin - (Improvements to suspend code for better DCON handling.) Modified: cpu/arm/olpc/1.75/fw.bth Modified: cpu/arm/olpc/1.75/fw.bth ============================================================================== --- cpu/arm/olpc/1.75/fw.bth Fri Sep 16 23:35:08 2011 (r2518) +++ cpu/arm/olpc/1.75/fw.bth Fri Sep 16 23:38:23 2011 (r2519) @@ -177,6 +177,9 @@ fload ${BP}/cpu/x86/pc/olpc/plot.fth fload ${BP}/cpu/arm/olpc/1.75/testinstructions.fth +fload ${BP}/cpu/arm/mmp2/dramrecal.fth +fload ${BP}/cpu/arm/mmp2/rtc.fth \ Internal RTC, used for wakeups + code halt ( -- ) wfi c; fload ${BP}/cpu/arm/olpc/1.75/switches.fth \ Lid and ebook switches

1 0

[commit] r2518 - cpu/arm/mmp2 cpu/arm/olpc/1.75 dev/olpc/dcon
by repository service Sept. 16, 2011

Sept. 16, 2011

Author: wmb Date: Fri Sep 16 23:35:08 2011 New Revision: 2518 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/2518 Log: OLPC XO-1.75 - Improvements to suspend code for better DCON handling. Modified: cpu/arm/mmp2/dramrecal.fth cpu/arm/olpc/1.75/devices.fth cpu/arm/olpc/1.75/lcd.fth dev/olpc/dcon/mmp2dcon.fth Modified: cpu/arm/mmp2/dramrecal.fth ============================================================================== --- cpu/arm/mmp2/dramrecal.fth Fri Sep 16 02:47:50 2011 (r2517) +++ cpu/arm/mmp2/dramrecal.fth Fri Sep 16 23:35:08 2011 (r2518) @@ -327,9 +327,9 @@ \ Security processor setup sleep-depth 4 >= if \ state at least POWER_MODE_CHIP_SLEEP (turn off most of SoC) -\ h# fe08.6000 0 mpmu! \ In SP, set AXISD, resvd, SLPEN, SPSD, DDRCORSD, APBSD resvd, VCXOSD + h# fe08.6000 0 mpmu! \ In SP, set AXISD, resvd, SLPEN, SPSD, DDRCORSD, APBSD resvd, VCXOSD \ We keep PMUM_BBDP (bit 25) off because that saves 60 mW - h# fc08.6000 0 mpmu! \ In SP, set AXISD, resvd, SLPEN, SPSD, DDRCORSD, APBSD resvd, VCXOSD +\ h# fc08.6000 0 mpmu! \ In SP, set AXISD, resvd, SLPEN, SPSD, DDRCORSD, APBSD resvd, VCXOSD then sleep-depth 3 = if \ state at least POWER_MODE_APPS_SLEEP (turn off slow IO) @@ -396,26 +396,28 @@ : keyboard-power-off ( -- ) d# 148 gpio-set ; : wlan-power-on ( -- ) d# 34 gpio-set ; : wlan-power-off ( -- ) d# 34 gpio-clr ; -: dcon-power-on ( -- ) 1 h# 26 ec-cmd-b! ; -: dcon-power-off ( -- ) 0 h# 26 ec-cmd-b! ; -h# ffff value sleep-mask +0 value sleep-mask : screen-off - sleep-mask 1 and if h# 12 " mode!" $call-screen then \ DCON power down - 0 h# 190 " lcd!" $call-screen - 0 h# 4c pmua! \ Kill the display clocks - saves 100 mW - \ 0 h# 54 pmua! \ Kill the SDIO 0 clocks - insignificant savings - \ 0 h# 58 pmua! \ Kill the SDIO 1 clocks - insignificant savings - sleep-mask 2 and if dcon-power-off then \ saves 80 mW - sleep-mask 4 and if keyboard-power-off then \ Should save about 17 mW - sleep-mask 8 and if wlan-power-off then \ saves 100 mW + sleep-mask 1 and if \ DCON power down + dcon-freeze + else + " dcon-suspend" $call-screen + then + " suspend" $call-screen + \ 0 h# 54 pmua! \ Kill the SDIO 0 clocks - insignificant savings + \ 0 h# 58 pmua! \ Kill the SDIO 1 clocks - insignificant savings + sleep-mask 2 and 0= if keyboard-power-off then \ Should save about 17 mW + sleep-mask 4 and 0= if wlan-power-off then \ saves 100 mW ; : screen-on ( -- ) - sleep-mask 8 and if wlan-power-on then - sleep-mask 4 and if keyboard-power-on then - sleep-mask 2 and if dcon-power-on d# 50 ms then \ saves 80 mW - h# 71b h# 4c pmua! - h# 8001100 h# 190 " lcd!" $call-screen - sleep-mask 1 and if h# 69 " mode!" $call-screen then \ DCON power up + sleep-mask 4 and 0= if wlan-power-on then + sleep-mask 2 and 0= if keyboard-power-on then + " resume" $call-screen + sleep-mask 1 and if \ DCON power up + dcon-unfreeze + else + " dcon-resume" $call-screen + then ; : stdin-idle-on ['] safe-idle to stdin-idle d# 15 enable-interrupt ; Modified: cpu/arm/olpc/1.75/devices.fth ============================================================================== --- cpu/arm/olpc/1.75/devices.fth Fri Sep 16 02:47:50 2011 (r2517) +++ cpu/arm/olpc/1.75/devices.fth Fri Sep 16 23:35:08 2011 (r2518) @@ -396,9 +396,6 @@ fload ${BP}/cpu/arm/mmp2/thermal.fth -fload ${BP}/cpu/arm/mmp2/dramrecal.fth -fload ${BP}/cpu/arm/mmp2/rtc.fth \ Internal RTC, used for wakeups - \ LICENSE_BEGIN \ Copyright (c) 2010 FirmWorks \ Modified: cpu/arm/olpc/1.75/lcd.fth ============================================================================== --- cpu/arm/olpc/1.75/lcd.fth Fri Sep 16 02:47:50 2011 (r2517) +++ cpu/arm/olpc/1.75/lcd.fth Fri Sep 16 23:35:08 2011 (r2518) @@ -2,12 +2,18 @@ : lcd@ ( offset -- l ) lcd-pa + io@ ; : lcd! ( l offset -- ) lcd-pa + io! ; -: init-lcd ( -- ) +: lcd-clock! ( n -- ) pmua-disp-clk-sel + h# 4c pmua! ; +: lcd-clocks-on ( -- ) \ Turn on clocks - h# 08 pmua-disp-clk-sel + h# 28284c io! - h# 09 pmua-disp-clk-sel + h# 28284c io! - h# 19 pmua-disp-clk-sel + h# 28284c io! - h# 1b pmua-disp-clk-sel + h# 28284c io! + h# 08 lcd-clock! + h# 09 lcd-clock! + h# 19 lcd-clock! + h# 1b lcd-clock! +; +: lcd-clocks-off ( -- ) 0 lcd-clock! ; + +: init-lcd ( -- ) + lcd-clocks-on 0 h# 190 lcd! \ Disable LCD DMA controller fb-mem-pa h# f4 lcd! \ Frame buffer area 0 @@ -205,3 +211,13 @@ 2drop ( ) flush-cursor-bits ( ) ; +0 value saved-mode +: suspend ( -- ) + h# 190 lcd@ to saved-mode + 0 h# 190 lcd! + lcd-clocks-off +; +: resume ( -- ) + lcd-clocks-on + saved-mode h# 190 lcd! +; Modified: dev/olpc/dcon/mmp2dcon.fth ============================================================================== --- dev/olpc/dcon/mmp2dcon.fth Fri Sep 16 02:47:50 2011 (r2517) +++ dev/olpc/dcon/mmp2dcon.fth Fri Sep 16 23:35:08 2011 (r2518) @@ -223,6 +223,22 @@ dcon-enable ( maybe-set-cmos ) \ dcon-enable leaves mode set to 69 - 40:antialias, 20:swizzle, 8:backlight on, 1:passthru off ; +: dcon-power-on ( -- ) 1 h# 26 ec-cmd-b! ; +: dcon-power-off ( -- ) 0 h# 26 ec-cmd-b! ; +0 value saved-dcon-mode +0 value saved-brightness +: dcon-suspend ( -- ) + bright@ to saved-brightness + mode@ to saved-dcon-mode + h# 12 mode! + dcon-power-off +; +: dcon-resume ( -- ) + dcon-power-on d# 50 ms + dcon-setup + saved-dcon-mode mode! + saved-brightness bright! +; \ LICENSE_BEGIN \ Copyright (c) 2010 FirmWorks

1 0

[commit] r2517 - in cpu/arm: mmp2 olpc/1.75
by repository service Sept. 15, 2011

Sept. 15, 2011

Author: wmb Date: Fri Sep 16 02:47:50 2011 New Revision: 2517 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/2517 Log: OLPC XO-1.75 - Checkpoint of suspend code - seems to be working for the most part. Modified: cpu/arm/mmp2/dramrecal.fth cpu/arm/mmp2/rtc.fth cpu/arm/olpc/1.75/prefw.bth Modified: cpu/arm/mmp2/dramrecal.fth ============================================================================== --- cpu/arm/mmp2/dramrecal.fth Fri Sep 16 02:46:57 2011 (r2516) +++ cpu/arm/mmp2/dramrecal.fth Fri Sep 16 02:47:50 2011 (r2517) @@ -97,11 +97,7 @@ mov r1, #0x80000000 \ PHY_SYNC_EN str r1, [r0, #0x240] \ PHY_CTRL14 - ldr r1, [r0, #0x230] \ PHY_CTRL13 - orr r1,r1,#0xf0000000 \ DLL_RESET_TIMER: 15*256 memory clocks -\ mov r1, #0xf0000000 \ DLL_RESET_TIMER: 15*256 memory clocks -\ orr r1, r1, #0x0040 \ DLL_PHSEL: 4 (45 + 4*3 = 57 degrees) - adjust? - str r1, [r0, #0x230] \ PHY_CTRL13 + \ There used to be code to set register 230, but Marvell says it is unnecessary. mov r1, #0x20000000 \ PHY_DLL_RESET str r1, [r0, #0x240] \ PHY_CTRL14 @@ -172,15 +168,6 @@ mov sp,r7 c; -: +apbc ( offset -- io-offset ) h# 01.5000 + ; -: +apmu ( offset -- io-offset ) h# 28.2800 + ; -: +mpmu ( offset -- io-offset ) h# 05.0000 + ; -: +scu ( offset -- io-offset ) h# 28.2c00 + ; -: +icu ( offset -- io-offset ) h# 28.2000 + ; - -: io-set ( mask offset -- ) dup io@ rot or swap io! ; -: io-clr ( mask offset -- ) dup io@ rot invert and swap io! ; - : apbc-clr-rst ( offset -- ) +apbc 4 swap io-clr ; : apbc-set-rst ( offset -- ) +apbc 4 swap io-set ; @@ -217,40 +204,40 @@ h# f0 h# 1c +scu io-clr \ SCU_MCB_CONF ; : disable-apmu-clks ( -- ) - h# 1b h# 54 +apmu io-clr \ PMUA_SDH0_CLK_RES_CTRL - h# 1b h# 58 +apmu io-clr \ PMUA_SDH1_CLK_RES_CTRL - h# 1b h# e8 +apmu io-clr \ PMUA_SDH2_CLK_RES_CTRL - h# 1b h# d4 +apmu io-clr \ PMUA_SMC_CLK_RES_CTRL - h# 3f h# 60 +apmu io-clr \ PMUA_NF_CLK_RES_CTRL - h# 3f h# d8 +apmu io-clr \ PMUA_MSPRO_CLK_RES_CTRL - XO does not use MSPRO - h# 12 h# 10c +apmu io-clr \ PMUA_AUDIO_CLK_RES_CTRL - h# 1fffd h# dc +apmu io-clr \ PMUA_GLB_CLK_RES_CTRL -\ 0 h# 68 +apmu io! \ PMUA_WTM_CLK_RES_CTRL - h# 9 h# 5c +apmu io-clr \ PMUA_USB_CLK_RES_CTRL + h# 1b h# 54 +pmua io-clr \ PMUA_SDH0_CLK_RES_CTRL + h# 1b h# 58 +pmua io-clr \ PMUA_SDH1_CLK_RES_CTRL + h# 1b h# e8 +pmua io-clr \ PMUA_SDH2_CLK_RES_CTRL + h# 1b h# d4 +pmua io-clr \ PMUA_SMC_CLK_RES_CTRL + h# 3f h# 60 +pmua io-clr \ PMUA_NF_CLK_RES_CTRL + h# 3f h# d8 +pmua io-clr \ PMUA_MSPRO_CLK_RES_CTRL - XO does not use MSPRO + h# 12 h# 10c +pmua io-clr \ PMUA_AUDIO_CLK_RES_CTRL + h# 1fffd h# dc +pmua io-clr \ PMUA_GLB_CLK_RES_CTRL +\ 0 h# 68 pmua! \ PMUA_WTM_CLK_RES_CTRL + h# 9 h# 5c +pmua io-clr \ PMUA_USB_CLK_RES_CTRL ; : enable-apmu-clks ( -- ) - h# 1b h# 54 +apmu io-set \ PMUA_SDH0_CLK_RES_CTRL - h# 1b h# 58 +apmu io-set \ PMUA_SDH1_CLK_RES_CTRL - h# 1b h# e8 +apmu io-set \ PMUA_SDH2_CLK_RES_CTRL - h# 1b h# d4 +apmu io-set \ PMUA_SMC_CLK_RES_CTRL \ ??? what is this and why is it on? - h# 3f h# 60 +apmu io-set \ PMUA_NF_CLK_RES_CTRL \ Should this be on? -\ h# 3f h# d8 +apmu io-set \ PMUA_MSPRO_CLK_RES_CTRL - h# 12 h# 10c +apmu io-set \ PMUA_AUDIO_CLK_RES_CTRL - h# 1fffd h# dc +apmu io-set \ PMUA_GLB_CLK_RES_CTRL -\ h# 1b h# 68 +apmu io! \ PMUA_WTM_CLK_RES_CTRL - h# 9 h# 5c +apmu io-set \ PMUA_USB_CLK_RES_CTRL + h# 1b h# 54 +pmua io-set \ PMUA_SDH0_CLK_RES_CTRL + h# 1b h# 58 +pmua io-set \ PMUA_SDH1_CLK_RES_CTRL + h# 1b h# e8 +pmua io-set \ PMUA_SDH2_CLK_RES_CTRL + h# 1b h# d4 +pmua io-set \ PMUA_SMC_CLK_RES_CTRL \ ??? what is this and why is it on? + h# 3f h# 60 +pmua io-set \ PMUA_NF_CLK_RES_CTRL \ Should this be on? +\ h# 3f h# d8 +pmua io-set \ PMUA_MSPRO_CLK_RES_CTRL + h# 12 h# 10c +pmua io-set \ PMUA_AUDIO_CLK_RES_CTRL + h# 1fffd h# dc +pmua io-set \ PMUA_GLB_CLK_RES_CTRL +\ h# 1b h# 68 pmua! \ PMUA_WTM_CLK_RES_CTRL + h# 9 h# 5c +pmua io-set \ PMUA_USB_CLK_RES_CTRL ; : disable-mpmu-clks ( -- ) - h# a010 h# 1024 +mpmu io! \ MPMU_ACGR + h# a010 h# 1024 mpmu! \ MPMU_ACGR - h# a010 h# 0024 +mpmu io! \ MPMU_CGR_SP ??? + h# a010 h# 0024 mpmu! \ MPMU_CGR_SP ??? \ h# 2000.0000 h# 414 +mpmu io-clr \ MPMU_PLL2_CTRL1 h# 8000.0000 h# 0040 +mpmu io-clr \ MPMU_ISCCR1 h# 8000.0000 h# 0044 +mpmu io-clr \ MPMU_ISCCR2 ; : enable-mpmu-clks ( -- ) - h# dffe.fffe h# 1024 +mpmu io! \ MPMU_ACGR + h# dffe.fffe h# 1024 mpmu! \ MPMU_ACGR \ h# 2000.0000 h# 0414 +mpmu io-set \ MPMU_PLL2_CTRL1 h# 8000.0000 h# 0040 +mpmu io-set \ MPMU_ISCCR1 h# 8000.0000 h# 0044 +mpmu io-set \ MPMU_ISCCR2 @@ -309,8 +296,8 @@ \ WAKEUP7: PMIC INT 5 value sleep-depth : setup-wakeup-sources ( -- mask ) - h# 0002.0094 h# 4c +mpmu io! \ RTC_ALARM, WAKEUP7, WAKEUP4, WAKEUP2 -\ h# 0002.0094 h# 104c +mpmu io! \ RTC_ALARM, WAKEUP7, WAKEUP4, WAKEUP2 ??? + h# 0002.0094 h# 4c mpmu! \ RTC_ALARM, WAKEUP7, WAKEUP4, WAKEUP2 +\ h# 0002.0094 h# 104c mpmu! \ RTC_ALARM, WAKEUP7, WAKEUP4, WAKEUP2 ??? h# ff08.7fff ( mask ) \ Enable all wakeup ports ; : block-irqs ( -- ) 1 h# 110 +icu io! ; @@ -319,8 +306,8 @@ 0 value save-idlecfg : restore-run-state ( -- ) - save-apcr h# 1000 +mpmu io! ( ) \ Restore APCR - save-idlecfg h# 18 +apmu io! ( ) \ Restore IDLE_CFG + save-apcr h# 1000 mpmu! ( ) \ Restore APCR + save-idlecfg h# 18 pmua! ( ) \ Restore IDLE_CFG ; \ Questions: \ Meaning of various connect type options @@ -340,21 +327,25 @@ \ Security processor setup sleep-depth 4 >= if \ state at least POWER_MODE_CHIP_SLEEP (turn off most of SoC) - h# fe08.6000 0 +mpmu io! \ In SP, set AXISD, resvd, SLPEN, SPSD, DDRCORSD, APBSD resvd, VCXOSD +\ h# fe08.6000 0 mpmu! \ In SP, set AXISD, resvd, SLPEN, SPSD, DDRCORSD, APBSD resvd, VCXOSD + \ We keep PMUM_BBDP (bit 25) off because that saves 60 mW + h# fc08.6000 0 mpmu! \ In SP, set AXISD, resvd, SLPEN, SPSD, DDRCORSD, APBSD resvd, VCXOSD then sleep-depth 3 = if \ state at least POWER_MODE_APPS_SLEEP (turn off slow IO) - h# de00.6000 0 +mpmu io! \ In SP, set AXISD, resvd, SPSD, DDRCORSD, APBSD resvd, +\ h# de00.6000 0 mpmu! \ In SP, set AXISD, resvd, SPSD, DDRCORSD, APBSD resvd, + \ We keep PMUM_BBDP (bit 25) off because that saves 60 mW + h# dc00.6000 0 mpmu! \ In SP, set AXISD, resvd, SPSD, DDRCORSD, APBSD resvd, then \ Linux value -\ h# 8030.0020 h# 14 +apmu io! \ IN PMUA_SP_IDLE_CFG, set , DIVIDER_RESET_EN, SP_DIS_MC_SW_REQ, SP_MC_WAKE_EN, TCM_STATE_RETAIN - h# 8020.0020 h# 14 +apmu io! \ IN PMUA_SP_IDLE_CFG, set , DIVIDER_RESET_EN, SP_DIS_MC_SW_REQ, TCM_STATE_RETAIN +\ h# 8030.0020 h# 14 pmua! \ IN PMUA_SP_IDLE_CFG, set , DIVIDER_RESET_EN, SP_DIS_MC_SW_REQ, SP_MC_WAKE_EN, TCM_STATE_RETAIN + h# 8020.0020 h# 14 pmua! \ IN PMUA_SP_IDLE_CFG, set , DIVIDER_RESET_EN, SP_DIS_MC_SW_REQ, TCM_STATE_RETAIN \ PJ4 setup - h# 1000 +mpmu io@ to save-apcr - h# 18 +apmu io@ to save-idlecfg + h# 1000 mpmu@ to save-apcr + h# 18 pmua@ to save-idlecfg save-apcr ( apcr ) h# ac08.0000 invert and ( apcr' ) \ Clear AXISD, SLPEN, DDRCORSD, APBSD, VCXOSD @@ -393,36 +384,69 @@ h# 0020.0000 or ( apcr idle' ) \ PJ_DIS_MC_SW_REQ - disable idle entry using software register bits h# 0010.0000 or ( apcr idle' ) \ PJ_MC_WAKE_EN - wake memory controller when core wakes - 0 h# b0 +apmu io! ( apcr idle ) \ PMUA_MC_HW_SLP_TYPE - self-refresh power down + 0 h# b0 pmua! ( apcr idle ) \ PMUA_MC_HW_SLP_TYPE - self-refresh power down - h# 18 +apmu io! ( apcr ) \ Set IDLE_CFG register - h# 1000 +mpmu io! ( ) \ Set APCR register + h# 18 pmua! ( apcr ) \ Set IDLE_CFG register + h# 1000 mpmu! ( ) \ Set APCR register \ end mmp2_pm_enter_lowpower_mode(state) ; -: screen-dark - " screen-ih iselect h# 12 mode! 0 h# 190 lcd! iunselect" eval \ Saves 1.05W - 0 h# 4c +apmu io! \ Kill the display clocks - saves 100 mW - \ 0 h# 54 +apmu io! \ Kill the SDIO 0 clocks - insignificant savings - \ 0 h# 58 +apmu io! \ Kill the SDIO 1 clocks - insignificant savings - d# 34 gpio-clr \ Kill the WLAN power - saves 100 mW +: keyboard-power-on ( -- ) d# 148 gpio-clr ; +: keyboard-power-off ( -- ) d# 148 gpio-set ; +: wlan-power-on ( -- ) d# 34 gpio-set ; +: wlan-power-off ( -- ) d# 34 gpio-clr ; +: dcon-power-on ( -- ) 1 h# 26 ec-cmd-b! ; +: dcon-power-off ( -- ) 0 h# 26 ec-cmd-b! ; +h# ffff value sleep-mask +: screen-off + sleep-mask 1 and if h# 12 " mode!" $call-screen then \ DCON power down + 0 h# 190 " lcd!" $call-screen + 0 h# 4c pmua! \ Kill the display clocks - saves 100 mW + \ 0 h# 54 pmua! \ Kill the SDIO 0 clocks - insignificant savings + \ 0 h# 58 pmua! \ Kill the SDIO 1 clocks - insignificant savings + sleep-mask 2 and if dcon-power-off then \ saves 80 mW + sleep-mask 4 and if keyboard-power-off then \ Should save about 17 mW + sleep-mask 8 and if wlan-power-off then \ saves 100 mW +; +: screen-on ( -- ) + sleep-mask 8 and if wlan-power-on then + sleep-mask 4 and if keyboard-power-on then + sleep-mask 2 and if dcon-power-on d# 50 ms then \ saves 80 mW + h# 71b h# 4c pmua! + h# 8001100 h# 190 " lcd!" $call-screen + sleep-mask 1 and if h# 69 " mode!" $call-screen then \ DCON power up ; + : stdin-idle-on ['] safe-idle to stdin-idle d# 15 enable-interrupt ; -: stdin-idle-off ['] noop to stdin-idle install-uart-io d# 15 disable-interrupt ; +: stdin-idle-off ['] noop to stdin-idle ( install-uart-io ) d# 15 disable-interrupt ; : timers-off ( -- ) 0 h# 14048 io! \ Disable interrupts from the tick timer 7 h# 1407c io! \ Clear any pending interrupts h# f disable-interrupt \ Block timer interrupt ; +: timers-on ( -- ) + 1 h# 14048 io! \ Enable interrupts from the tick timer + 7 h# 1407c io! \ Clear any pending interrupts + h# f enable-interrupt \ Unblock timer interrupt + reschedule-tick +; + +: power-islands-off ( -- ) + 0 h# 10c pmua! \ Turn off audio power island +; +: power-islands-on ( -- ) + h# 712 h# 10c pmua! \ Turn on audio power island +; : str ( -- ) + disable-interrupts timers-off - screen-dark + screen-off stdin-idle-off - 5 h# 38 +mpmu io! \ Use 32 kHz clock instead of VCXO for slow clock + 5 h# 38 mpmu! \ Use 32 kHz clock instead of VCXO for slow clock \ OLPC: Unmask main PMU interrupt - don't know if this is necessary \ h# 400 h# 174 +icu io-clr @@ -435,13 +459,13 @@ setup-sleep-state - h# 000c.0000 h# 8c +apmu io-set \ Power down CoreSight SRAM + h# 000c.0000 h# 8c +pmua io-set \ Power down CoreSight SRAM \ TODO - need to power down sram/l2$ \ mmp2_cpu_disable_l2(0); \ outer_cache.flush_range(0, -1ul); - \ TODO - handle low power island? + power-islands-off disable-clks @@ -454,11 +478,15 @@ enable-clks - ." STR out" cr + power-islands-on + \ mmp2_cpu_enable_l2(0); \ idle_cfg &= (~PMUA_MOH_SRAM_PWRDWN); stdin-idle-on + screen-on + timers-on + enable-interrupts ; \ LICENSE_BEGIN Modified: cpu/arm/mmp2/rtc.fth ============================================================================== --- cpu/arm/mmp2/rtc.fth Fri Sep 16 02:46:57 2011 (r2516) +++ cpu/arm/mmp2/rtc.fth Fri Sep 16 02:47:50 2011 (r2517) @@ -10,8 +10,8 @@ : int5-status@ ( -- value ) h# 28.2154 io@ ; : enable-rtc ( -- ) h# 81 h# 01.5000 io! ; : enable-rtc-wakeup ( -- ) - h# 5.004c io@ h# 2.0010 or h# 5.004c io! - h# 5.104c io@ h# 2.0010 or h# 5.104c io! + h# 004c mpmu@ h# 2.0010 or h# 004c mpmu! + h# 104c mpmu@ h# 2.0010 or h# 104c mpmu! ; : soc-rtc@ ( offset -- value ) h# 01.0000 + io@ ; : soc-rtc! ( value offset -- value ) h# 01.0000 + io! ; Modified: cpu/arm/olpc/1.75/prefw.bth ============================================================================== --- cpu/arm/olpc/1.75/prefw.bth Fri Sep 16 02:46:57 2011 (r2516) +++ cpu/arm/olpc/1.75/prefw.bth Fri Sep 16 02:47:50 2011 (r2517) @@ -168,6 +168,22 @@ : io! ( value offset -- ) +io l! ; : io@ ( offset -- value ) +io l@ ; +: +apbc ( offset -- io-offset ) h# 01.5000 + ; +: +pmua ( offset -- io-offset ) h# 28.2800 + ; +: +mpmu ( offset -- io-offset ) h# 05.0000 + ; +: +scu ( offset -- io-offset ) h# 28.2c00 + ; +: +icu ( offset -- io-offset ) h# 28.2000 + ; + +: io-set ( mask offset -- ) dup io@ rot or swap io! ; +: io-clr ( mask offset -- ) dup io@ rot invert and swap io! ; + + +: mpmu@ ( offset -- l ) +mpmu io@ ; +: mpmu! ( l offset -- ) +mpmu io! ; + +: pmua@ ( offset -- l ) +pmua io@ ; +: pmua! ( l offset -- ) +pmua io! ; + fload ${BP}/cpu/arm/mmp2/hash.fth \ Hashes - SHA1, SHA-256, MD5 fload ${BP}/cpu/x86/pc/olpc/crypto.fth \ Cryptographic image validation fload ${BP}/cpu/x86/pc/olpc/lzip.fth \ Access zip images from memory

1 0

[commit] r2516 - cpu/arm/olpc/1.75
by repository service Sept. 15, 2011

Sept. 15, 2011

Author: wmb Date: Fri Sep 16 02:46:57 2011 New Revision: 2516 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/2516 Log: OLPC XO-1.75 - Better management of ACK line to EC when not ready to receive events - part 2. Modified: cpu/arm/olpc/1.75/fw.bth Modified: cpu/arm/olpc/1.75/fw.bth ============================================================================== --- cpu/arm/olpc/1.75/fw.bth Fri Sep 16 02:43:49 2011 (r2515) +++ cpu/arm/olpc/1.75/fw.bth Fri Sep 16 02:46:57 2011 (r2516) @@ -32,8 +32,12 @@ unload-crypto " /usb" " reset-usb" execute-device-method drop ; +: quiesce ( -- ) + usb-quiet + close-ec +; -' usb-quiet to linux-hook +' quiesce to linux-hook d# 9999 to arm-linux-machine-type \ Marvell Jasper

1 0