OpenBIOS October 2007

openbios@openbios.org

11 participants
63 discussions

[PATCH] Read size of installed memory from LinuxBIOS table
by Jens Freimann 12 Oct '07

12 Oct '07

Hi, this patch makes OFW read the size of installed RAM from the LinuxBIOS table instead of fetching it from mem-info-pa. With kind regards, Jens

3 3

r673 - dev/olpc/kb3700
by svn＠openbios.org 12 Oct '07

12 Oct '07

Author: wmb Date: 2007-10-12 00:15:10 +0200 (Fri, 12 Oct 2007) New Revision: 673 Modified: dev/olpc/kb3700/ecio.fth Log: OLPC ecio.fth - added new command ec-indexed-io-off for later use when we pull the plug on indexed I/O. Modified: dev/olpc/kb3700/ecio.fth =================================================================== --- dev/olpc/kb3700/ecio.fth 2007-10-10 00:18:33 UTC (rev 672) +++ dev/olpc/kb3700/ecio.fth 2007-10-11 22:15:10 UTC (rev 673) @@ -154,6 +154,8 @@ : write-protect-fw ( -- ) h# 29 ec-cmd ; : ebook-mode? ( -- b ) h# 2a ec-cmd-b@ ; +: ec-indexed-io-off ( -- ) h# fe95 ec@ h# 40 invert and h# fe95 ec! ; + 0 [if] \ EC mailbox access words

1 0

[PATCH] Create device tree by parsing LinuxBIOS device tree
by Jens Freimann 10 Oct '07

10 Oct '07

Hi, this patch adds support to create the OFW device tree from the LinuxBIOS device tree, so it doesn't need to do all the probing etc. again. With kind regards, Jens

1 0

r672 - ofw/fs/jffs2
by svn＠openbios.org 10 Oct '07

10 Oct '07

Author: wmb Date: 2007-10-10 02:18:33 +0200 (Wed, 10 Oct 2007) New Revision: 672 Modified: ofw/fs/jffs2/jffs2.fth Log: JFFS2 reader - When a page read occurs, don't abort and don't discard previously-read good pages within that erase block. Modified: ofw/fs/jffs2/jffs2.fth =================================================================== --- ofw/fs/jffs2/jffs2.fth 2007-10-10 00:15:02 UTC (rev 671) +++ ofw/fs/jffs2/jffs2.fth 2007-10-10 00:18:33 UTC (rev 672) @@ -118,21 +118,29 @@ : page>eblock ( page# -- eblock# ) pages/eblock / ; -1 ( instance ) value have-eblock# \ For avoiding redundant reads -: read-pages ( page# #pages -- error? ) +: (read-pages) ( page# #pages -- #read ) \ Partial reads invalidate the cache dup pages/eblock <> if -1 to have-eblock# then ( page# #pages ) - tuck block-buf -rot " read-blocks" $call-parent ( #pages #read ) - <> + block-buf -rot " read-blocks" $call-parent ( #read ) ; +: read-pages ( page# #pages -- error? ) tuck (read-pages) <> ; : read-eblock ( eblock# -- ) - dup have-eblock# <> if ( eblock# ) - to have-eblock# ( ) - have-eblock# eblock>page pages/eblock read-pages - abort" jffs2: bad read" - else - drop - then ( ) + dup have-eblock# <> if ( eblock# ) + to have-eblock# ( ) + have-eblock# eblock>page pages/eblock (read-pages) ( npages ) + dup pages/eblock <> if ( npages ) + ." JFFS2: bad read - eblock# " have-eblock# .x ." page " dup .x cr + /page * ( block-offset ) + dup block-buf + ( block-offset adr ) + /eblock rot - ( adr erase-length ) + h# ff fill ( ) + else ( npages ) + drop ( ) + then ( ) + else ( eblock# ) + drop ( ) + then ( ) ; 0 ( instance ) value sumsize

1 0

r671 - cpu/x86/pc/olpc
by svn＠openbios.org 10 Oct '07

10 Oct '07

Author: wmb Date: 2007-10-10 02:15:02 +0200 (Wed, 10 Oct 2007) New Revision: 671 Modified: cpu/x86/pc/olpc/security.fth Log: OLPC security - When processing signature blobs, search for sig01: lines, ignoring others, thus paving the way for new formats in the future. Modified: cpu/x86/pc/olpc/security.fth =================================================================== --- cpu/x86/pc/olpc/security.fth 2007-10-10 00:10:18 UTC (rev 670) +++ cpu/x86/pc/olpc/security.fth 2007-10-10 00:15:02 UTC (rev 671) @@ -163,14 +163,35 @@ " Signature valid" ?lease-debug-cr then ; -: sha-valid? ( data$ sig01$ -- okay? ) " sha256" invalid? 0= ; -: fw-valid? ( data$ 2*sig$ -- okay? ) - 2swap 2>r ( 2*sig$ r: data$ ) - newline left-parse-string ( rmd-sig$ sha-sig$ r: data$ ) - 2r@ 2swap sha-valid? 0= if ( rmd-sig$ r: data$ ) + +\ Look for a line that starts with "sig01: " +: next-sig01$ ( sig$ -- true | rem$ sig01$ false ) + begin dup while ( rem$ ) + newline left-parse-string ( rem$' line$ ) + 2dup 7 min " sig01: " $= if ( rem$ line$ ) + false exit + then ( rem$ line$ ) + 2drop ( rem$ ) + repeat ( rem$ ) + 2drop true +; + +\ Find a sig01: line and check its sha256/rsa signature +: sha-valid? ( data$ sig01$ -- okay? ) + next-sig01$ if 2drop false exit then ( data$ rem$ sig01$ ) + 2nip " sha256" invalid? 0= +; + +\ Find two sig01: lines, the first with sha256 and the seconcd with rmd160, +\ and check their signatures +: fw-valid? ( data$ sig$ -- okay? ) + 2swap 2>r ( sig$ r: data$ ) + next-sig01$ if 2r> 2drop false exit then ( rem$ sig01$ ) + 2r@ 2swap sha-valid? 0= if ( rem$ r: data$ ) 2r> 4drop false exit - then ( rmd-sig$ r: data$ ) - 2r> 2swap " rmd160" invalid? 0= + then ( rmd-sig$ r: data$ ) + next-sig01$ if 2r> 2drop false exit then ( rem$ sig01$ ) + 2nip 2r> 2swap " rmd160" invalid? 0= ; \ earliest is the earliest acceptable date value (in seconds).

1 0

r670 - cpu/x86/pc/olpc
by svn＠openbios.org 10 Oct '07

10 Oct '07

Author: wmb Date: 2007-10-10 02:10:18 +0200 (Wed, 10 Oct 2007) New Revision: 670 Added: cpu/x86/pc/olpc/timertest.fth Modified: cpu/x86/pc/olpc/devices.fth Log: OLPC Trac #4050 - added selftest /isa/timer to catch broken 14 MHz clocks. Modified: cpu/x86/pc/olpc/devices.fth =================================================================== --- cpu/x86/pc/olpc/devices.fth 2007-10-10 00:08:32 UTC (rev 669) +++ cpu/x86/pc/olpc/devices.fth 2007-10-10 00:10:18 UTC (rev 670) @@ -64,6 +64,8 @@ [then] +fload ${BP}/cpu/x86/pc/olpc/timertest.fth \ Selftest for PIT timer + 1 [if] dev /interrupt-controller h# 20 to vector-base0 @@ -111,6 +113,9 @@ [then] fload ${BP}/dev/pci/isaall.fth +\ We don't need a serial selftest because the serial port is internal only +\ and the selftest turns off the diag device +dev /serial warning @ warning off : selftest false ; warning ! device-end devalias com1 /isa/serial@i3f8:115200 devalias mouse /isa/8042/mouse devalias d disk Added: cpu/x86/pc/olpc/timertest.fth =================================================================== --- cpu/x86/pc/olpc/timertest.fth (rev 0) +++ cpu/x86/pc/olpc/timertest.fth 2007-10-10 00:10:18 UTC (rev 670) @@ -0,0 +1,61 @@ +purpose: Selftest for ISA timer +\ See license at end of file + +dev /isa/timer + +: wait-rollover ( -- error? ) + get-msecs d# 200 + begin ( time-limit ) + 0 count@ 1 = if drop false exit then + dup get-msecs - ( time-limit diff ) + 0<= until + drop true +; +: selftest ( -- error? ) + 0 count@ 0= if + 1 ms + 0 count@ 0= if + ." The ISA PIT timer is not ticking." cr + ." Possible cause: bad 14 MHz clock to CS5536 chip." cr + true exit + then + then + + wait-rollover if + ." The ISA PIT timer did not reach a count of 1." cr + true exit + then + + 1 ms + + 0 count@ d# 10000 d# 11000 between 0= if + ." The ISA PIT timer is ticking at the wrong rate." cr + true exit + then + + false +; + +device-end +\ LICENSE_BEGIN +\ Copyright (c) 2007 FirmWorks +\ +\ Permission is hereby granted, free of charge, to any person obtaining +\ a copy of this software and associated documentation files (the +\ "Software"), to deal in the Software without restriction, including +\ without limitation the rights to use, copy, modify, merge, publish, +\ distribute, sublicense, and/or sell copies of the Software, and to +\ permit persons to whom the Software is furnished to do so, subject to +\ the following conditions: +\ +\ The above copyright notice and this permission notice shall be +\ included in all copies or substantial portions of the Software. +\ +\ THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, +\ EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF +\ MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND +\ NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE +\ LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION +\ OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION +\ WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. +\ +\ LICENSE_END

1 0

r669 - ofw/core
by svn＠openbios.org 10 Oct '07

10 Oct '07

Author: wmb Date: 2007-10-10 02:08:32 +0200 (Wed, 10 Oct 2007) New Revision: 669 Modified: ofw/core/ofwcore.fth Log: test-all - If an individual test fails, highlight (red letters) the error message and wait 10 seconds so the user can see it. Modified: ofw/core/ofwcore.fth =================================================================== --- ofw/core/ofwcore.fth 2007-10-10 00:07:12 UTC (rev 668) +++ ofw/core/ofwcore.fth 2007-10-10 00:08:32 UTC (rev 669) @@ -3211,7 +3211,12 @@ method-name 2@ current-device ( method-adr,len phandle ) execute-phandle-method drop ( result ) - ?dup if ??cr ." Selftest failed. Return code = " .d cr then + ?dup if + red-letters + ??cr ." Selftest failed. Return code = " .d cr + d# 10000 ms + black-letters + then then ( ) ;

1 0

r668 - cpu/x86/pc/olpc
by svn＠openbios.org 10 Oct '07

10 Oct '07

Author: wmb Date: 2007-10-10 02:07:12 +0200 (Wed, 10 Oct 2007) New Revision: 668 Modified: cpu/x86/pc/olpc/resume.bth Log: OLPC resume - Fixed that "reset-smbus-bitbang" code that unwedges the SMBUS, but left it commented out, because the kernel does it too. Modified: cpu/x86/pc/olpc/resume.bth =================================================================== --- cpu/x86/pc/olpc/resume.bth 2007-10-05 18:36:19 UTC (rev 667) +++ cpu/x86/pc/olpc/resume.bth 2007-10-10 00:07:12 UTC (rev 668) @@ -340,33 +340,32 @@ [ifdef] reset-smbus-bitbang \ GPIO15 is SMB_DATA \ GPIO14 is SMB_CLOCK + h# c000 # ax mov \ Mask to set SMB_DATA and SMB_CLOCK + h# 1000 # dx mov ax dx out \ Set output values to high + h# 1004 # dx mov ax dx out \ Set pins to output h# c0000000 # ax mov \ Mask to clear SMB_DATA and SMB_CLOCK h# 1010 # dx mov ax dx out \ Deselect OUT AUX1 h# 1014 # dx mov ax dx out \ Deselect OUT AUX2 h# 1034 # dx mov ax dx out \ Deselect IN AUX1 - h# c000 # ax mov \ Mask to set SMB_DATA and SMB_CLOCK - h# 1000 # dx mov ax dx out \ Set output values to high - h# 1004 # dx mov ax dx out \ Set pins to output - h# 1000 # dx mov \ Output value register - - 8 # cx mov \ Generate 8 low pulses on SMB_CLOCK + d# 16 # cx mov \ Generate 8 low pulses on SMB_CLOCK begin \ 5 uS delay (slightly longer for GX) rdtsc ax bx mov d# 5 d# 500 * # bx add begin rdtsc bx ax sub 0>= until + h# 1000 # dx mov \ Output value register h# 40000000 # ax mov ax dx out \ Clear SMB_CLOCK \ 5 uS delay (slightly longer for GX) rdtsc ax bx mov d# 5 d# 500 * # bx add begin rdtsc bx ax sub 0>= until + h# 1000 # dx mov \ Output value register h# 4000 # ax mov ax dx out \ Set SMB_CLOCK - loopa - \ 50 uS delay (slightly longer for GX) + \ 5 uS delay (slightly longer for GX) rdtsc ax bx mov d# 5 d# 500 * # bx add begin rdtsc bx ax sub 0>= until [then]

1 0

Instantly switching between Windows and Linux
by Marcos Reyes 10 Oct '07

10 Oct '07

I would like to know if it would be possible, now or in the future, to instantly switch between two running operating systems on one machine. For example by dividing the RAM on the machine to be dedicated to each OS, and having a buffer for device states. It would be really cool, for example, to launch a shortcut from within windows that launches a program in Linux. Though some multitasking would be lost, as processes in Windows would be suspended. Or maybe not, with multi-processor systems? I don't know anything about this... just a dream. Would be great to migrate people to Linux. M

1 0

[PATCH] Fix Linux/x86 load without virtual-mode -- take 2
by Torsten Duwe 08 Oct '07

08 Oct '07

> I'll have to see yet what's wrong with earlyprintk. Linux fixups didn't get called because there was no way the Linux ELF kernel could be recognised as such. So the fix to the fix is a one-liner plus a comment. The recogniser needs further refinement, but on top of LinuxBIOSv2 this boots Linux at least in qemu and on my real machine. Please apply! Signed-off-by: Torsten Duwe <duwe(a)lst.de>

1 0