December 2007 - OpenBIOS - mail.coreboot.org

r753 - cpu/x86/pc/olpc
by svn＠openbios.org Dec. 7, 2007

Dec. 7, 2007

Author: wmb Date: 2007-12-07 10:14:42 +0100 (Fri, 07 Dec 2007) New Revision: 753 Modified: cpu/x86/pc/olpc/versions.fth Log: Q2D07 - svn 741, plus cherry-picked 749 750 752 753 Modified: cpu/x86/pc/olpc/versions.fth =================================================================== --- cpu/x86/pc/olpc/versions.fth 2007-12-07 09:03:55 UTC (rev 752) +++ cpu/x86/pc/olpc/versions.fth 2007-12-07 09:14:42 UTC (rev 753) @@ -2,7 +2,7 @@ \ The overall firmware revision macro: FW_MAJOR D -macro: FW_MINOR 06 +macro: FW_MINOR 07 \ The EC microcode macro: EC_VERSION test4e

1 0

r752 - ofw/fs
by svn＠openbios.org Dec. 7, 2007

Dec. 7, 2007

Author: wmb Date: 2007-12-07 10:03:55 +0100 (Fri, 07 Dec 2007) New Revision: 752 Modified: ofw/fs/unixtime.fth Log: OLPC trac 5391 - Don't crash if >unix-seconds is presented with an invalid month. Modified: ofw/fs/unixtime.fth =================================================================== --- ofw/fs/unixtime.fth 2007-12-05 23:12:36 UTC (rev 751) +++ ofw/fs/unixtime.fth 2007-12-07 09:03:55 UTC (rev 752) @@ -31,7 +31,8 @@ : >unix-seconds ( s m h d m y -- seconds ) \ since 1970 d# 1970 - 4 /mod [ d# 365 4 * 1+ ] literal * ( s m h d m yrs days ) swap d# 365 * + ( s m h d m days ) - swap 1- 0 ?do i days/month + c@ + loop ( s m h d days ) + swap 1 max d# 12 min ( s m h d days m' ) + 1- 0 ?do i days/month + c@ + loop ( s m h d days ) + 1- ( s m h days ) d# 24 * + d# 60 * + d# 60 * + ;

1 0

r751 - dev/usb2/device/storage
by svn＠openbios.org Dec. 5, 2007

Dec. 5, 2007

Author: wmb Date: 2007-12-06 00:12:36 +0100 (Thu, 06 Dec 2007) New Revision: 751 Modified: dev/usb2/device/storage/scsidisk.fth Log: OLPC trac 5354 - retry read-capacity command for USB mass storage devices that need extra startup time. Modified: dev/usb2/device/storage/scsidisk.fth =================================================================== --- dev/usb2/device/storage/scsidisk.fth 2007-12-05 03:44:34 UTC (rev 750) +++ dev/usb2/device/storage/scsidisk.fth 2007-12-05 23:12:36 UTC (rev 751) @@ -67,11 +67,18 @@ \ The SCSI-2 standard requires disk devices to implement \ the "read capacity" command. - 8 read-capacity-cmd 0a short-data-command if ( ) - true - else ( adr ) - dup 4 + 4c@ swap 4c@ 1+ false - then + \ Retry it a few times just in case that helps + d# 20 0 do + 8 read-capacity-cmd 0a short-data-command 0= if ( ) + dup 4 + 4c@ swap 4c@ 1+ false + unloop exit + then + d# 200 ms + loop + + \ If it fails, we just guess. Some devices violate the spec and + \ fail to implement read_capacity + d# 512 h# ffffffff false ; external

1 0

r750 - cpu/x86/pc/olpc
by svn＠openbios.org Dec. 4, 2007

Dec. 4, 2007

Author: wmb Date: 2007-12-05 04:44:34 +0100 (Wed, 05 Dec 2007) New Revision: 750 Modified: cpu/x86/pc/olpc/versions.fth Log: Q2D06 - build from svn 741 + 749 + 750 Modified: cpu/x86/pc/olpc/versions.fth =================================================================== --- cpu/x86/pc/olpc/versions.fth 2007-12-05 01:57:25 UTC (rev 749) +++ cpu/x86/pc/olpc/versions.fth 2007-12-05 03:44:34 UTC (rev 750) @@ -2,7 +2,7 @@ \ The overall firmware revision macro: FW_MAJOR D -macro: FW_MINOR 05 +macro: FW_MINOR 06 \ The EC microcode macro: EC_VERSION test4e @@ -16,7 +16,7 @@ \ The wireless LAN module firmware macro: WLAN_RPM ${WLAN_VERSION}-1.olpc1 -macro: WLAN_VERSION 5.110.20.p1 +macro: WLAN_VERSION 5.110.20.p42 \ The bios_verify image macro: CRYPTO_VERSION 0.2

1 0

r749 - ofw/fs/jffs2
by svn＠openbios.org Dec. 4, 2007

Dec. 4, 2007

Author: wmb Date: 2007-12-05 02:57:25 +0100 (Wed, 05 Dec 2007) New Revision: 749 Modified: ofw/fs/jffs2/jffs2.fth Log: OLPC trac 5322 - If the first dirent in block n+1 had the same name and pino as the last dirent in block n, but has an earlier version, it incorrectly superseded the last dirent. This was a side effect of an earlier optimization to eliminate the storing of junk dirents. Modified: ofw/fs/jffs2/jffs2.fth =================================================================== --- ofw/fs/jffs2/jffs2.fth 2007-12-03 05:27:55 UTC (rev 748) +++ ofw/fs/jffs2/jffs2.fth 2007-12-05 01:57:25 UTC (rev 749) @@ -836,6 +836,7 @@ inodes 'next-inode ! pages/chip 0 do i page>eblock to the-eblock# + 0 prev-name c! i h# 3fff and 0= if scan-callout then i no-summary? if i possible-nodes? if i scan-raw-nodes then

1 0

r748 - ofw/inet
by svn＠openbios.org Dec. 2, 2007

Dec. 2, 2007

Author: wmb Date: 2007-12-03 06:27:55 +0100 (Mon, 03 Dec 2007) New Revision: 748 Modified: ofw/inet/telnetd.fth Log: Telnetd - fixed incompatibility with the Windows telnet client - looping on option negotiation and doubling carriage returns. Modified: ofw/inet/telnetd.fth =================================================================== --- ofw/inet/telnetd.fth 2007-12-02 10:35:56 UTC (rev 747) +++ ofw/inet/telnetd.fth 2007-12-03 05:27:55 UTC (rev 748) @@ -55,7 +55,9 @@ : will ( rem$ -- rem$' ) next-cmd-byte " WILL" .got dup case - 0 of send-do endof +\ Since we have already sent "do binary", there is no need to re-ack it +\ 0 of send-do endof + 0 of drop endof \ Suppress go-ahead \ Since we have already sent "do suppressGA", there is no need to re-ack it \ 3 of send-do endof \ Suppress go-ahead @@ -125,16 +127,26 @@ endcase ; +\ Remove the linefeed in a cr-lf pair. +\ The Windows telnet client sends CR-LF per the Telnet NVT spec. +\ The Linux telnet client just sends CR. +0 value last-was-cr? + +: do-lf ( adr len -- adr' len' ) + last-was-cr? if swallow else 1 /string then +; + : process-escapes ( adr len -- len' ) - over swap ( adr rem$ ) - begin dup while ( adr rem$ ) - over c@ #iac = if ( adr rem$ ) - do-command ( adr rem$' ) - else ( adr rem$ ) - 1 /string ( adr rem$' ) - then ( adr rem$ ) - repeat ( adr end-adr 0 ) - drop swap - ( len' ) + over swap ( adr rem$ ) + begin dup while ( adr rem$ ) + over c@ dup >r case ( adr rem$ r: char ) + #iac of do-command endof ( adr rem$' r: char ) + linefeed of do-lf endof ( adr rem$' r: char ) + ( adr rem$ char ) >r 1 /string r> ( adr rem$' r: char ) + endcase ( adr rem$ r: char ) + r> carret = to last-was-cr? ( adr rem$ ) + repeat ( adr end-adr 0 ) + drop swap - ( len' ) ; : read ( adr len -- actual ) over swap (read) ( adr actual ) @@ -174,13 +186,22 @@ begin get-msecs over d# 300 + - 0< while - the-byte 1 (read) case ( msecs [byte] ) - 1 of drop get-msecs the-byte 1 process-escapes drop endof - -1 of drop false exit endof - endcase + the-byte h# 10 (read) dup 0< if ( msecs count ) + \ Bail out if the connection closed + -1 = if drop false exit then + + \ The other alternative is -2, meaning no bytes available. + \ In that case, we'll eventually time out. + else ( msecs count ) + \ Keep advancing the timeout while bytes are still coming in + the-byte swap process-escapes drop ( msecs ) + drop get-msecs + then repeat drop + false to last-was-cr? + true ; : close ( -- ) ;

1 0

r747 - dev/usb2/device/keyboard
by svn＠openbios.org Dec. 2, 2007

Dec. 2, 2007

Author: wmb Date: 2007-12-02 11:35:56 +0100 (Sun, 02 Dec 2007) New Revision: 747 Modified: dev/usb2/device/keyboard/request.fth Log: USB keyboard driver - Use control-set instead of control-set-nostat because keyboard devices appear to generate a status phase. Modified: dev/usb2/device/keyboard/request.fth =================================================================== --- dev/usb2/device/keyboard/request.fth 2007-12-02 10:31:17 UTC (rev 746) +++ dev/usb2/device/keyboard/request.fth 2007-12-02 10:35:56 UTC (rev 747) @@ -38,12 +38,12 @@ : set-boot-protocol ( -- error? ) 0 0 my-address ( interface ) 0 DR_HIDD DR_OUT or SET_PROTOCOL - control-set-nostat + control-set ; : set-idle ( ms -- error? ) >r 0 0 my-address ( interface ) r> 4 / 8 << ( 4ms ) DR_HIDD DR_OUT or SET_IDLE - control-set-nostat + control-set ; \ Key modifiers @@ -63,7 +63,7 @@ : (set-leds) ( led -- ) led-buf c! led-buf /led-buf my-address ( interface ) REPORT_OUT DR_HIDD DR_OUT or SET_REPORT - control-set-nostat drop + control-set drop ; : set-leds ( led-mask -- ) dup to led-state (set-leds) ; : toggle-leds ( led-mask -- ) led-state xor set-leds ;

1 0

r746 - in cpu/x86/pc: biosload olpc
by svn＠openbios.org Dec. 2, 2007

Dec. 2, 2007

Author: wmb Date: 2007-12-02 11:31:17 +0100 (Sun, 02 Dec 2007) New Revision: 746 Modified: cpu/x86/pc/biosload/ofw.bth cpu/x86/pc/olpc/loaddropins.fth Log: USB - Load the usbkbd driver only for the keyboard subclass; don't load it for the generic hid subclass. Modified: cpu/x86/pc/biosload/ofw.bth =================================================================== --- cpu/x86/pc/biosload/ofw.bth 2007-12-02 10:27:07 UTC (rev 745) +++ cpu/x86/pc/biosload/ofw.bth 2007-12-02 10:31:17 UTC (rev 746) @@ -80,9 +80,9 @@ " ${BP}/dev/usb2/hcd/ohci/build/ohci.fc" " class0c0310" $add-deflated-dropin " ${BP}/dev/usb2/hcd/ehci/build/ehci.fc" " class0c0320" $add-deflated-dropin - " ${BP}/dev/usb2/device/hub/build/hub.fc" " usb,class9" $add-dropin - " ${BP}/dev/usb2/device/keyboard/build/usbkbd.fc" " usb,class3" $add-dropin -\ " ${BP}/dev/usb2/device/net/build/usbnet.fc" " usbnet" $add-deflated-dropin + " ${BP}/dev/usb2/device/hub/build/hub.fc" " usb,class9" $add-dropin + " ${BP}/dev/usb2/device/keyboard/build/usbkbd.fc" " usb,class3,1" $add-dropin + " ${BP}/dev/usb2/device/net/build/usbnet.fc" " usbnet" $add-deflated-dropin " ${BP}/dev/usb2/device/serial/build/usbserial.fc" " usbserial" $add-deflated-dropin " ${BP}/dev/usb2/device/storage/build/usbstorage.fc" " usbstorage" $add-deflated-dropin Modified: cpu/x86/pc/olpc/loaddropins.fth =================================================================== --- cpu/x86/pc/olpc/loaddropins.fth 2007-12-02 10:27:07 UTC (rev 745) +++ cpu/x86/pc/olpc/loaddropins.fth 2007-12-02 10:31:17 UTC (rev 746) @@ -8,7 +8,7 @@ " ${BP}/dev/usb2/hcd/ehci/build/ehci.fc" " class0c0320" $add-deflated-dropin " ${BP}/dev/usb2/device/hub/build/hub.fc" " usb,class9" $add-dropin " ${BP}/dev/usb2/device/net/build/usbnet.fc" " usbnet" $add-deflated-dropin - " ${BP}/dev/usb2/device/keyboard/build/usbkbd.fc" " usb,class3" $add-dropin + " ${BP}/dev/usb2/device/keyboard/build/usbkbd.fc" " usb,class3,1" $add-dropin " ${BP}/dev/usb2/device/serial/build/usbserial.fc" " usbserial" $add-deflated-dropin " ${BP}/dev/usb2/device/storage/build/usbstorage.fc" " usbstorage" $add-deflated-dropin " ${BP}/dev/usb2/device/wlan/build/usb8388.fc" " usb1286,2001" $add-deflated-dropin

1 0

r745 - in dev/usb2/hcd: . ehci uhci
by svn＠openbios.org Dec. 2, 2007

Dec. 2, 2007

Author: wmb Date: 2007-12-02 11:27:07 +0100 (Sun, 02 Dec 2007) New Revision: 745 Modified: dev/usb2/hcd/ehci/probe.fth dev/usb2/hcd/hcd.fth dev/usb2/hcd/uhci/bulk.fth dev/usb2/hcd/uhci/probe.fth dev/usb2/hcd/uhci/uhci.fth Log: USB UHCI driver - Several fixes for the UHCI (USB 1.1) driver for compatibility with Intel ICH7 and some full-speed devices. Modified: dev/usb2/hcd/ehci/probe.fth =================================================================== --- dev/usb2/hcd/ehci/probe.fth 2007-12-02 02:43:57 UTC (rev 744) +++ dev/usb2/hcd/ehci/probe.fth 2007-12-02 10:27:07 UTC (rev 745) @@ -40,6 +40,29 @@ then drop ; +: grab-controller ( -- error? ) + hccparams@ 8 rshift h# ff and dup if ( config-adr ) + dup my-l@ h# 10001 = if ( config-adr ) + h# 100.0000 over my-l! ( config-adr ) \ Ask for it + true ( config-adr error? ) + d# 100 0 do ( config-adr error? ) + over my-l@ h# 101.0000 and h# 100.0000 = if + \ Turn off SMIs in Legacy Support Extended CSR + h# e000.0000 h# 6c my-l! ( config-adr error? ) + 0 my-l@ h# 27cc8086 = if + h# ffff.0000 h# 70 my-l! \ Clear EHCI Intel special SMIs + then + 0= leave ( config-adr error?' ) + then ( config-adr error? ) + d# 10 ms ( config-adr error? ) + loop ( config-adr error? ) + nip exit + then ( config-adr ) + then ( config-adr ) + drop ( ) + false +; + external : power-usb-ports ( -- ) ; @@ -64,8 +87,14 @@ parse-my-args open-count 0= if map-regs + alloc-dma-buf first-open? if false to first-open? + grab-controller if + ." Can't take control of EHCI from underlying BIOS" cr + free-dma-buf unmap-regs + false exit + then 0 ehci-reg@ h# ff and to op-reg-offset reset-usb init-ehci-regs @@ -74,8 +103,6 @@ init-struct init-extra then - alloc-dma-buf - probe-root-hub then open-count 1+ to open-count Modified: dev/usb2/hcd/hcd.fth =================================================================== --- dev/usb2/hcd/hcd.fth 2007-12-02 02:43:57 UTC (rev 744) +++ dev/usb2/hcd/hcd.fth 2007-12-02 10:27:07 UTC (rev 745) @@ -60,6 +60,9 @@ : my-w@ ( offset -- w ) my-space + " config-w@" $call-parent ; : my-w! ( w offset -- ) my-space + " config-w!" $call-parent ; +: my-l@ ( offset -- l ) my-space + " config-l@" $call-parent ; +: my-l! ( l offset -- ) my-space + " config-l!" $call-parent ; + : map-in ( phys.lo,md,hi len -- vaddr ) " map-in" $call-parent ; : map-out ( vaddr size -- ) " map-out" $call-parent ; Modified: dev/usb2/hcd/uhci/bulk.fth =================================================================== --- dev/usb2/hcd/uhci/bulk.fth 2007-12-02 02:43:57 UTC (rev 744) +++ dev/usb2/hcd/uhci/bulk.fth 2007-12-02 10:27:07 UTC (rev 745) @@ -103,21 +103,24 @@ bulk-in-qh my-speed insert-bulk-qh ; +: bulk-in-actual ( -- actual usberr ) + bulk-in-td bulk-in-qh >qh-#tds l@ get-actual ( actual ) + usb-error ( actual usberr ) + bulk-in-td bulk-in-qh >qh-#tds l@ fixup-bulk-in-data +; + : bulk-in? ( -- actual usberr ) bulk-in-qh 0= if 0 USB_ERR_INV_OP exit then clear-usb-error process-hc-status - bulk-in-qh dup sync-qhtds - qh-done? if - bulk-in-td bulk-in-qh >qh-#tds l@ get-actual ( actual ) - usb-error ( actual usberr ) - bulk-in-td bulk-in-qh >qh-#tds l@ fixup-bulk-in-data - else + bulk-in-qh dup sync-qhtds ( bulk-in-qh ) + qh-done? if ( ) + bulk-in-actual ( actual usberr ) + else ( ) bulk-in-qh dup >hcqh-elem le-l@ ( qh elem ) 1 ms over sync-qhtds ( qh elem ) swap >hcqh-elem le-l@ = if \ No movement in the past ms - bulk-in-td bulk-in-qh >qh-#tds l@ get-actual ( actual ) - usb-error ( actual usberr ) + bulk-in-actual ( actual usberr ) bulk-in-td bulk-in-qh >qh-#tds l@ fixup-bulk-in-data else \ It may not be done yet 0 usb-error ( actual usberr ) @@ -130,14 +133,16 @@ headers : restart-bulk-in-td ( td -- ) - begin ?dup while - dup >hctd-token dup le-l@ TD_TOKEN_DATA1 invert and - bulk-in-data@ toggle-bulk-in-data or swap le-l! - dup >hctd-stat dup le-l@ - TD_STAT_ANY_ERROR TD_ACTUAL_MASK or invert and - TD_STAT_ACTIVE or swap le-l! - >td-next l@ - repeat + begin ?dup while ( td ) + toggle-bulk-in-data ( td ) + dup >hctd-token ( td &token ) + dup le-l@ TD_TOKEN_DATA1 invert and ( td &token value ) + bulk-in-data@ or swap le-l! ( td ) + dup >hctd-stat dup le-l@ ( td &stat value ) + TD_STAT_ANY_ERROR TD_ACTUAL_MASK or invert and ( td &stat value' ) + TD_STAT_ACTIVE or swap le-l! ( td ) + >td-next l@ ( td' ) + repeat ( ) ; external Modified: dev/usb2/hcd/uhci/probe.fth =================================================================== --- dev/usb2/hcd/uhci/probe.fth 2007-12-02 02:43:57 UTC (rev 744) +++ dev/usb2/hcd/uhci/probe.fth 2007-12-02 10:27:07 UTC (rev 745) @@ -4,17 +4,20 @@ hex headers +\ We mustn't wait more than 3 ms between releasing the reset and enabling +\ the port to begin the SOF stream, otherwise some devices (e.g. pl2303) +\ will go into suspend state and then not respond to set-address. : reset-root-hub-port ( port -- ) - dup >r portsc@ fff5 and 200 or r@ portsc! \ Reset port - 10 ms - r@ portsc@ fff5 and 200 invert and r@ portsc! - 10 ms - r@ portsc@ fff5 and 4 or r@ portsc! \ Enable port - 20 ms - r@ portsc@ a or r> portsc! \ Clear status + dup >r portsc@ h# 20e invert and ( value r: port ) \ Clear reset, enable, status + dup h# 200 or r@ portsc! ( value r: port ) \ Reset port + d# 30 ms ( value r: port ) \ > 10 ms - reset time + dup r@ portsc! ( value r: port ) \ Release reset + 1 ms ( value r: port ) \ > 5.3 uS - reconnect time + h# e or r> portsc! ( ) \ Enable port and clear status ; : probe-root-hub-port ( port -- ) + dup reset-root-hub-port dup portsc@ 1 and 0= if drop exit then \ No device-connected ok-to-add-device? 0= if drop exit then \ Can't add another device @@ -39,7 +42,6 @@ 2 0 do i portsc@ h# a and if i rm-obsolete-children \ Remove obsolete device nodes - i reset-root-hub-port i ['] probe-root-hub-port catch if drop ." Failed to probe root port " i .d cr then @@ -57,6 +59,7 @@ map-regs first-open? if false to first-open? + ?disable-smis reset-usb init-struct init-lists @@ -73,7 +76,7 @@ : close ( -- ) open-count 1- to open-count end-extra - open-count 0= if free-dma-buf unmap-regs then + open-count 0= if free-dma-buf then \ Don't unmap ; \ LICENSE_BEGIN Modified: dev/usb2/hcd/uhci/uhci.fth =================================================================== --- dev/usb2/hcd/uhci/uhci.fth 2007-12-02 02:43:57 UTC (rev 744) +++ dev/usb2/hcd/uhci/uhci.fth 2007-12-02 10:27:07 UTC (rev 745) @@ -52,6 +52,10 @@ : portsc@ ( port -- data ) 2* 10 + uhci-w@ ; : portsc! ( data port -- ) 2* 10 + uhci-w! ; +: ?disable-smis ( -- ) + 0 my-l@ h# 27c88086 = if h# af00 h# 80 my-w! then +; + : reset-usb ( -- ) uhci-reg dup 0= if map-regs then 4 usbcmd! \ Global reset

1 0

r744 - dev/usb2/hcd/ehci
by svn＠openbios.org Dec. 1, 2007

Dec. 1, 2007

Author: wmb Date: 2007-12-02 03:43:57 +0100 (Sun, 02 Dec 2007) New Revision: 744 Modified: dev/usb2/hcd/ehci/ehci.fth Log: USB EHCI driver - extended a setting time delay from 1 ms to 3 ms because 1 ms is empirically too short for some devices on systems that don't have individual port power controls. Modified: dev/usb2/hcd/ehci/ehci.fth =================================================================== --- dev/usb2/hcd/ehci/ehci.fth 2007-12-02 02:41:50 UTC (rev 743) +++ dev/usb2/hcd/ehci/ehci.fth 2007-12-02 02:43:57 UTC (rev 744) @@ -154,7 +154,7 @@ : claim-ownership ( -- ) 1 cfgflag! \ Claim ownership to all ports - 1 ms + 3 ms \ Give devices time to settle \ Power on ports if necessary hcsparams@ h# 10 and if

1 0