Menu
▾
▴
colinux-devel — Cooperative Linux development mailing list
You can subscribe to this list here.
| 2004 |
Jan
(64) |
Feb
(530) |
Mar
(266) |
Apr
(580) |
May
(360) |
Jun
(161) |
Jul
(185) |
Aug
(164) |
Sep
(123) |
Oct
(160) |
Nov
(59) |
Dec
(84) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2005 |
Jan
(156) |
Feb
(95) |
Mar
(124) |
Apr
(81) |
May
(79) |
Jun
(179) |
Jul
(35) |
Aug
(64) |
Sep
(56) |
Oct
(57) |
Nov
(18) |
Dec
(41) |
| 2006 |
Jan
(65) |
Feb
(37) |
Mar
(59) |
Apr
(73) |
May
(65) |
Jun
(27) |
Jul
(54) |
Aug
(76) |
Sep
(103) |
Oct
(23) |
Nov
(45) |
Dec
(29) |
| 2007 |
Jan
(41) |
Feb
(47) |
Mar
(61) |
Apr
(24) |
May
(14) |
Jun
(6) |
Jul
(23) |
Aug
(30) |
Sep
(16) |
Oct
(9) |
Nov
(53) |
Dec
(36) |
| 2008 |
Jan
(19) |
Feb
(49) |
Mar
(74) |
Apr
(21) |
May
(24) |
Jun
(5) |
Jul
(9) |
Aug
(53) |
Sep
(26) |
Oct
(23) |
Nov
(32) |
Dec
(19) |
| 2009 |
Jan
(47) |
Feb
(49) |
Mar
(39) |
Apr
(61) |
May
(28) |
Jun
(19) |
Jul
(12) |
Aug
(10) |
Sep
(31) |
Oct
(16) |
Nov
(60) |
Dec
(26) |
| 2010 |
Jan
(17) |
Feb
(9) |
Mar
(32) |
Apr
(11) |
May
(24) |
Jun
(33) |
Jul
(5) |
Aug
(2) |
Sep
(7) |
Oct
(8) |
Nov
(17) |
Dec
(7) |
| 2011 |
Jan
(12) |
Feb
(16) |
Mar
(2) |
Apr
(12) |
May
(5) |
Jun
(10) |
Jul
(3) |
Aug
(3) |
Sep
(2) |
Oct
(1) |
Nov
(17) |
Dec
(1) |
| 2012 |
Jan
(9) |
Feb
(9) |
Mar
(8) |
Apr
(4) |
May
(2) |
Jun
(1) |
Jul
(4) |
Aug
(8) |
Sep
(11) |
Oct
(1) |
Nov
(2) |
Dec
(2) |
| 2013 |
Jan
|
Feb
(7) |
Mar
(4) |
Apr
(10) |
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(5) |
Dec
(3) |
| 2016 |
Jan
|
Feb
(2) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2023 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
(1) |
|
From: Digital I. Inc. <ok...@di...> - 2004-02-27 11:17:30
|
Virtual Device Drivers
For next major release (version 1.0?) , we need more virtual device drivers, dont
you?. Your comment I hope.
--- Okajima.
----------------
- Virtual Sound driver
very easy.
- Virtual Mouse driver
easy. but it should be "seamless" like when you
installed VMware tools to VMware.
- Virtual Printer driver
print documents from coLinux through Windows.
this is easy if you rasterize rightly.
so if you can print on linux logically,
( solving font, cups/ghostscript settings...)
you can print on coLinux physically.
- Virtual Frame buffer
Performance is the key. BitBlt all frame is easy, but slow.
It must bitblt only a region which is updated. How?
And second point is, is this really necessary?
If you use only X, cygwin/X is enough.
- Virtual FB / How to know an updated region.
1. Installing a special driver to X/SDL.
If you use only a popular graphics library, this way
is the best. but does not work for minor graphics lib
and some games which access FB directly.
2. Using MMU Dirty bits.
if your apps modify FB only "horizontally", this works
very well. but for example, you draw a vertical line,
it does not work.
|
|
From: Digital I. Inc. <ok...@di...> - 2004-02-27 11:17:28
|
Performance Tuning : rough idea
Okay, coLinux is running on many PCs. A lot of problems, but running. Then it is
time to start a discussion about how to tweak coLinux.
I think coLinux has problems mainly in three criteria.
- X
- File access
- Memory shortage
Before starting, it is wise that we discuss many possibilities from many aspects.
Your idea, comments, anything is welcome.
Of course, other than them, we have many problems and many points to improve.
Your opinion is definitely welcome.
---- Okajima.
-------------
- X
I think it is not so slow even at this moment. But you can improve it by using
virtual inter process communication (IPC) between coLinux and Windows through
a passage, insted of using virtual TCP link currently used.
Some people say that making X run on coLinux natively, it speeds up. but I am
doubtful about this. because if do so, you can not use Windows 2D graphics
accelerator driver.
- File Access
This is slow. and I think direct sector access to NTFS partition though NT
sector read driver would help.
- Memory shortage
coLinux has only very little memory. There is a lot of ways to solve this,
and your idea is welcome. I think creating virtual UMA(using main DRAM as VRAM -
used mainly with on-board VGA.) device on NT is one way. It's memory is not
going to be swapped, probably.
Another idea is, using this. http://linuxcompressed.sourceforge.net/
Second, coLinux uses memory doubly. for example, it caches sector data doubly.
I mean, NT caches them. and coLinux caches same data.
Someday coLinux can run multiple instance, and the problem will become more
serious. because, there are many same ELF, same .so and same something....
between VMs.
XEN ( http://www.cl.cam.ac.uk/Research/SRG/netos/xen/ ) is working on this problem.
In next version ( XEN 1.3 ), they have a plan to use shared ELF and .so between
VMs. And even it shares chache buffer.
And Jeff Dike, the author of UML, has same plan.
|
|
From: <bv...@pr...> - 2004-02-27 09:49:41
|
Problem as described in http://sourceforge.net/mailarchive/forum.php?thread_id=3913390&forum_id=37489 but on Win2K SP4. Exception: STATUS_ACCESS_VIOLATION at eip=10007512 eax=00000198 ebx=00405044 ecx=6110B0C8 edx=00000000 esi=00000000 edi=00405048 ebp=00E2EF74 esp=00E2EF44 program=C:\colinux\colinux-bridged-net-daemon.exe cs=001B ds=0023 es=0023 fs=0038 gs=0000 ss=0023 Stack trace: Frame Function Args 00E2EF74 10007512 (00000000, 00E2EFAC, 00403880, 00000000) 00E2EFA4 61003D64 (00000000, 00000000, 00000000, 00000000) 00E2FFA4 61003D0E (00000000, 00000000, 00000000, 00000000) End of stack trace -- Vadim |
|
From: Marko B. <mar...@al...> - 2004-02-27 08:53:15
|
Hi! I have installed coLinux on a Win2k (SP4) and a WinXP (SP1) box. Winpcap is the latest 3.1 beta. Bridged networking is working on Wink2k, but it crashes on WinXP. Here the colinux-bridged-net-daemon.exe.stackdump file: Exception: STATUS_ACCESS_VIOLATION at eip=10007B52 eax=00000194 ebx=00405044 ecx=6110B0C8 edx=00000000 esi=00000000 edi=00405048 ebp=0117EF74 esp=0117EF44 program=C:\colinux\colinux-bridged-net-daemon.exe cs=001B ds=0023 es=0023 fs=0038 gs=0000 ss=0023 Stack trace: Frame Function Args 0117EF74 10007B52 (00000000, 0117EFAC, 00403880, 00000000) 0117EFA4 61003D64 (00000000, 00000000, 00000000, 00000000) 0117FFA4 61003D0E (00000000, 00000000, 00000000, 00000000) End of stack trace packet.dll and wpcap.dll in coLinux distribution are v3.0.0.19 (from Winpcap 3.0), Winpcap DLLs have version 3.1.0.20. I have tried installing Winpcap 3.0, but the same thing happens. I've seen a message on the coLinux forum that an older version of Winpcap (with DLL version 3.0.0.16) works on XP, but I still have to try that. Any idea where the bug is? P.S. Great work with coLinux! It's great! -- Marko ICQ: 5990814 I'm not under the alkafluence of inkahol that some thinkle peep I am. It's just the drunker I sit here the longer I get. |
|
From: SofExtR4/<So...@So...> - 2004-02-27 02:54:41
|
Incident Information:- Originator: col...@li... Recipients: col...@li... Subject: [coLinux-devel] 1GB, 2GB, and 4GB blank filesystems. Generic Alert: Exception was 'Corrupt File' scanning attachment 'empty-4gb.bz2' |
|
From: Pat E. <pat...@ma...> - 2004-02-27 02:37:47
|
Seeing as people have been having trouble with creating blank block devices, here are 5. the 3 'empty' are just 0s. the 2 swap ones have had mkswap done on them already. TAKE NOTE, they do decompress to 256MB, 512MB, 1GB, 2GB, 4GB.... Pat Erley |
|
From: Alex <ro...@ya...> - 2004-02-27 00:48:17
|
Hi,
I checked reiserfs sources to see where it reports error during boot
and its stuck on this piece of code:
if ( blk_size[MAJOR(dev)][MINOR(dev)] <
sb_block_count(rs)*(sb_blocksize(rs)>>10) ) {
printk("Filesystem on %s cannot be mounted because it is bigger than
the device\n", kdevname(dev));
printk("You may need to run fsck or increase size of your LVM
partition\n");
printk("Or may be you forgot to reboot after fdisk when it told you to\n");
return NULL;
}
I guess blk_size[MAJOR(dev)][MINOR(dev)] should return real size of
codb0 block device but here with such config line in default.colinux.xml
<block_device index="0" path="\Device\HarddiskVolume2"
enabled="true"></block_device>
it just return 0. Can it be fixed so it return real size ? Please, give
me a hand with these, or point where in colinux size of block device
calculations should be. I know i can just get rid of that condition in
reiserfs and everything probably will work, but thats not right, you
know ;).
--
With best regards, Alex
|
|
From: <ch...@to...> - 2004-02-26 18:35:56
|
> On Thursday 26 February 2004 17:48, ch...@to... wrote: >> I am working on a dual boot setup native or coLinux. I have the fstab >> situation worked out with so sym link tricks. The problem I am having >> now >> is with networking. I want to use eth0 with one config for native boot >> (same IP address the windows setup uses to comunicate with outside) and >> when under coLinux use a different interface on a private inside lan and >> have window NAT it out. I will then work out the route issue in linux. >> >> I tried >> >> <network index="1" type="tap" name="TAP"></network> >> >> but when it booted coLinux ifconfig showed that etho was up and ther was >> no eth1 >> >> I realize that only 1 device is visable in each boot mode. >> >> I also saw this in the change log >> >> I also prepared the ground for supporting more than one network >> device >> (i.e, eth1, eth2, etc.), but it is not yet completed. >> >> is this the problem or is there some other workaround anyone can think >> of? >> I want to be able to use static addresses and I only have one address >> for >> comunicating with the outside world that is why I want to have windows >> NAT >> the address when in coLinux and when Native use the same address as in >> the >> windows config. > If you want a nice solution to switching files around dependent on your > environment check out http://hprofile.sourceforge.net > Essentially what it allows for is determining a profile to be used through > kernel parameters or checking for devices that would exist in one > environment, but not another etc (you need to customise that through > scripts). It switches to a runlevel accordingly and takes care of all of > the > symlinks for you. Works fantastic. Looks interesting. do the kernel parameters pass through coLinux daemon fine? I saw something a while back that someone couldn't get ro to pass. right now I am using the fact that topologilinux uses an initrd to boot so I have links to the real devices on the initrd and the coLinux devices in the directory that is the mountpoint for the initrd. That seems to work good for switching / and swap between /dev/cobd? and /dev/loop? I'll definatly look into hprofile further. chris > >> >> I am taking notes as I do this and should have a good HowTo to post on >> the >> wiki (which is looking very nice, very quickly) >> >> I am using Topologilinux which runs off of loopback files on the windows >> partition. Booting both ways so far but I want to get networking and >> maby >> floppy and cdrom access (I set them up in the rootfs but need to figure >> out the colinux.xml setup for them. >> >> chris >> >> >> ------------------------------------------------------- >> SF.Net is sponsored by: Speed Start Your Linux Apps Now. >> Build and deploy apps & Web services for Linux with >> a free DVD software kit from IBM. Click Now! >> http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click >> _______________________________________________ >> coLinux-devel mailing list >> coL...@li... >> https://lists.sourceforge.net/lists/listinfo/colinux-devel > |
|
From: <ch...@to...> - 2004-02-26 18:25:17
|
That is what I was getting. The linux side seemed to ignore the index. > On Thu, Feb 26, 2004 at 06:48:33PM +0100, ch...@to... wrote: > >> I also saw this in the change log >> >> I also prepared the ground for supporting more than one network >> device >> (i.e, eth1, eth2, etc.), but it is not yet completed. >> >> is this the problem or is there some other workaround anyone can think >> of? > > It is really incomplete, you shouldn't play with it. > The Linux side currently ignores the index. > > -- > Dan Aloni > da...@gm... > |
|
From: Sean B. <sea...@so...> - 2004-02-26 18:23:08
|
On Thursday 26 February 2004 17:48, ch...@to... wrote: > I am working on a dual boot setup native or coLinux. I have the fstab > situation worked out with so sym link tricks. The problem I am having now > is with networking. I want to use eth0 with one config for native boot > (same IP address the windows setup uses to comunicate with outside) and > when under coLinux use a different interface on a private inside lan and > have window NAT it out. I will then work out the route issue in linux. > > I tried > > <network index="1" type="tap" name="TAP"></network> > > but when it booted coLinux ifconfig showed that etho was up and ther was > no eth1 > > I realize that only 1 device is visable in each boot mode. > > I also saw this in the change log > > I also prepared the ground for supporting more than one network > device > (i.e, eth1, eth2, etc.), but it is not yet completed. > > is this the problem or is there some other workaround anyone can think of? > I want to be able to use static addresses and I only have one address for > comunicating with the outside world that is why I want to have windows NAT > the address when in coLinux and when Native use the same address as in the > windows config. If you want a nice solution to switching files around dependent on your environment check out http://hprofile.sourceforge.net Essentially what it allows for is determining a profile to be used through kernel parameters or checking for devices that would exist in one environment, but not another etc (you need to customise that through scripts). It switches to a runlevel accordingly and takes care of all of the symlinks for you. Works fantastic. > > I am taking notes as I do this and should have a good HowTo to post on the > wiki (which is looking very nice, very quickly) > > I am using Topologilinux which runs off of loopback files on the windows > partition. Booting both ways so far but I want to get networking and maby > floppy and cdrom access (I set them up in the rootfs but need to figure > out the colinux.xml setup for them. > > chris > > > ------------------------------------------------------- > SF.Net is sponsored by: Speed Start Your Linux Apps Now. > Build and deploy apps & Web services for Linux with > a free DVD software kit from IBM. Click Now! > http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click > _______________________________________________ > coLinux-devel mailing list > coL...@li... > https://lists.sourceforge.net/lists/listinfo/colinux-devel |
|
From: Dan A. <da...@gm...> - 2004-02-26 18:07:46
|
On Thu, Feb 26, 2004 at 06:48:33PM +0100, ch...@to... wrote: > I also saw this in the change log > > I also prepared the ground for supporting more than one network > device > (i.e, eth1, eth2, etc.), but it is not yet completed. > > is this the problem or is there some other workaround anyone can think of? It is really incomplete, you shouldn't play with it. The Linux side currently ignores the index. -- Dan Aloni da...@gm... |
|
From: Pat E. <pat...@ma...> - 2004-02-26 17:59:01
|
On Thu, 26 Feb 2004 17:28:42 +0000 Sean Brook <sea...@so...> wrote: > When building a colinux kernel which options could be used and which options > are a very bad idea or just will not work? > > Some are obvious, like specific hardware support. But can I change the > processor in Processor type and features->Processor Family? > this is a VERY VERY arguable point right now. there have been recent changes that should/may have fixed that problem, but most people(including myself) will say that stick with i386 til more testing can be done. Pat Erley |
|
From: <ch...@to...> - 2004-02-26 17:58:34
|
I am working on a dual boot setup native or coLinux. I have the fstab
situation worked out with so sym link tricks. The problem I am having now
is with networking. I want to use eth0 with one config for native boot
(same IP address the windows setup uses to comunicate with outside) and
when under coLinux use a different interface on a private inside lan and
have window NAT it out. I will then work out the route issue in linux.
I tried
<network index="1" type="tap" name="TAP"></network>
but when it booted coLinux ifconfig showed that etho was up and ther was
no eth1
I realize that only 1 device is visable in each boot mode.
I also saw this in the change log
I also prepared the ground for supporting more than one network
device
(i.e, eth1, eth2, etc.), but it is not yet completed.
is this the problem or is there some other workaround anyone can think of?
I want to be able to use static addresses and I only have one address for
comunicating with the outside world that is why I want to have windows NAT
the address when in coLinux and when Native use the same address as in the
windows config.
I am taking notes as I do this and should have a good HowTo to post on the
wiki (which is looking very nice, very quickly)
I am using Topologilinux which runs off of loopback files on the windows
partition. Booting both ways so far but I want to get networking and maby
floppy and cdrom access (I set them up in the rootfs but need to figure
out the colinux.xml setup for them.
chris
|
|
From: Dan A. <da...@gm...> - 2004-02-26 17:53:16
|
On Thu, Feb 26, 2004 at 09:34:21PM +0600, Dmitriy Kazimirow wrote: > +++ colinux-20040225/src/Makefile 2004-02-25 08:13:22.000000000 +0600 > @@ -81,7 +81,7 @@ > > HOST_API_CFLAGS=-DCO_HOST_API > > -AR=i586-pc-cygwin-ar > +AR=i686-pc-cygwin-ar > >[snip] > > ifeq ($(COLINUX_ARCH),i386) > -CROSS_TARGET=i586-pc-cygwin- > +CROSS_TARGET=i686-pc-cygwin- > else > $(error Other Windows architectures not supported yet!) > endif > diff -r -u colinux-20040225.k6/src/colinux/os/winnt/build/Makefile colinux-20040225/src/colinux/os/winnt/build/Makefile > --- colinux-20040225.k6/src/colinux/os/winnt/build/Makefile 2004-02-26 03:29:09.109092800 +0600 > +++ colinux-20040225/src/colinux/os/winnt/build/Makefile 2004-02-25 05:21:30.000000000 +0600 > @@ -24,7 +24,7 @@ > > CLEAN_FILES += $(BUILD_PATH)/driver.base.tmp > $(BUILD_PATH)/driver.base.tmp: $(BUILD_PATH)/driver.o > - i586-pc-cygwin-gcc -Wl,--base-file,$@ \ > + i686-pc-cygwin-gcc -Wl,--base-file,$@ \ > -Wl,--entry,_DriverEntry@8 \ > -nostartfiles -nostdlib \ > -o junk.tmp $^ -lntoskrnl -lhal -lgcc > @@ -32,12 +32,12 @@ The prefix of the cygwin's tool chain executables denotes about the platform of the compiler and not about the code it produces, e.g, you can compile an i686 code by passing -march=i686 to i586-pc-cygwin-gcc and compile an i586 code by passing -march=i586 to i686-pc-cygwin-gcc. AFAIK, the default compilation target is i386 unless specified otherwise. Both the Windows driver and userspace supplied with the binary distribution of coLinux are compiled without any '-march='-style parameters meaning that they are compatible with i386. So, the meaningful part in your patch is the where you replace fx* instructions with f*. I plan for future versions of coLinux to detect whether fxsave can be used or not and then choose the right passage page code. Another remark is that the vmlinux file distributed with coLinux versions up to 0.5.4 was compiled with -march=i386. In 0.5.4, it was compiled with -march=i686. This change will be reverted in the next versions. -- Dan Aloni da...@gm... |
|
From: Sean B. <sea...@so...> - 2004-02-26 17:40:42
|
When building a colinux kernel which options could be used and which options are a very bad idea or just will not work? Some are obvious, like specific hardware support. But can I change the processor in Processor type and features->Processor Family? |
|
From: Bart K. <bar...@su...> - 2004-02-26 17:09:31
|
> I used cygwin X. > > On coLinux (Debian), use dselect to find xterm and install it or do an > "apt-get install xterm" instead. > > From cygwin, use "xhost +" to allow any connection, and that is okay behind > a firewall. I copied all relevant files from cygwin for the XServer and deleted the rest of cygwin leaving me with only 6 Mb of DLL's to run the XWin server OUTSIDE cygwin (packed about 2.5 MB) ;) I zipped all of these and put them online, and put a bat file to startup the server. Tested with KDE and Enlightment, works fine a compact XServer for the till coLinux supports X native. http://speeldoos.showcase.surfnet.nl/startx.zip regards, Bart |
|
From: Ballard J. <sac...@ho...> - 2004-02-26 17:08:21
|
I put the patch on Sourceforge.net that implements the NT/Linux console. Here is a beta release that works with 0.5.4: ftp://ftp.drunkencodepoets.com:821/colinux-console.0.5.4.bjh.1.beta.exe.b= z2 |
|
From: Dan A. <da...@gm...> - 2004-02-26 16:04:57
|
On Thu, Feb 26, 2004 at 09:34:21PM +0600, Dmitriy Kazimirow wrote: > Welcome all > I tried using 0.5.3 and 0.5.4 on old machine with amd k6-3+ CPU > 0.5.3 runs ok but 0.5.4 crash my win2k > (I tried only Debian 1gb imaged,updated to 3.0r1 using Debian CDs) > After investing this issue I think I found the problem: > 1.almost all is cross-compiled with i686-pc-cygwin > 2.context switch code in src\colinux\arch\i386\passage.c use > fxsave/fxrestor(which just not avaible on old CPUs)(fsave/frestor is > avaible but not save SSE context but SSE is not exist on k6 anyway) > > here is patch for build 0.5.4 which will work on old CPUs > (possible problem is that SSE state will not be saved when patched > version runs on modern CPUs,better use some #define to detect minimal > target CPU) Did you actually test it? The patch below is a reversed one, BTW. I'll try to see how we can merge it in. -- Dan Aloni da...@gm... |
|
From: Dmitriy K. <dm...@ka...> - 2004-02-26 15:48:11
|
Welcome all
I tried using 0.5.3 and 0.5.4 on old machine with amd k6-3+ CPU
0.5.3 runs ok but 0.5.4 crash my win2k
(I tried only Debian 1gb imaged,updated to 3.0r1 using Debian CDs)
After investing this issue I think I found the problem:
1.almost all is cross-compiled with i686-pc-cygwin
2.context switch code in src\colinux\arch\i386\passage.c use
fxsave/fxrestor(which just not avaible on old CPUs)(fsave/frestor is
avaible but not save SSE context but SSE is not exist on k6 anyway)
here is patch for build 0.5.4 which will work on old CPUs
(possible problem is that SSE state will not be saved when patched
version runs on modern CPUs,better use some #define to detect minimal
target CPU)
diff -r -u colinux-20040225.k6/src/Makefile colinux-20040225/src/Makefile
--- colinux-20040225.k6/src/Makefile 2004-02-26 03:28:39.496512000 +0600
+++ colinux-20040225/src/Makefile 2004-02-25 08:13:22.000000000 +0600
@@ -81,7 +81,7 @@
HOST_API_CFLAGS=-DCO_HOST_API
-AR=i586-pc-cygwin-ar
+AR=i686-pc-cygwin-ar
TOOL_USER_CC=$(CC) $(USER_CFLAGS) $(CO_INCLUDES) -c $^ -o $@
TOOL_USER_GPP=$(GPP) $(USER_CFLAGS) $(CO_INCLUDES) -c $^ -o $@
diff -r -u colinux-20040225.k6/src/colinux/arch/i386/passage.c colinux-20040225/src/colinux/arch/i386/passage.c
--- colinux-20040225.k6/src/colinux/arch/i386/passage.c 2004-02-26 21:24:34.790468800 +0600
+++ colinux-20040225/src/colinux/arch/i386/passage.c 2004-02-22 04:55:14.000000000 +0600
@@ -65,8 +65,9 @@
/* save return address */
" movl %ebx, 0x38(%ebp)" "\n"
-/* " fxsave 0x70(%ebp)" "\n" */
- " fsave 0x70(%ebp)" "\n"
+
+/* save FPU / MMX / SSE state */
+ " fxsave 0x70(%ebp)" "\n"
/* Put the virtual address of the passage page in EBX */
" movl %ecx, %ebx" "\n"
@@ -141,6 +142,7 @@
" movl %ebx, 0x0C(%ebp)" "\n"
" movl %cs, %ebx" "\n"
" movl %ebx, 0x04(%ebp)" "\n"
+
/* be on the safe side and nullify the segment registers */
" movl $0, %ebx" "\n"
" movl %ebx, %fs" "\n"
@@ -312,7 +314,7 @@
"1:" "\n"
/* restore FPU / MMX / SSE state */
- " frstor 0x70(%ebp)" "\n"
+ " fxrstor 0x70(%ebp)" "\n"
/* get old ESP in EAX */
" lss 0x40(%ebp), %eax" "\n"
diff -r -u colinux-20040225.k6/src/colinux/os/winnt/Makefile colinux-20040225/src/colinux/os/winnt/Makefile
--- colinux-20040225.k6/src/colinux/os/winnt/Makefile 2004-02-26 03:58:59.884100800 +0600
+++ colinux-20040225/src/colinux/os/winnt/Makefile 2004-02-05 03:32:54.000000000 +0600
@@ -7,7 +7,7 @@
#-----------------------------------------------------------------
ifeq ($(COLINUX_ARCH),i386)
-CROSS_TARGET=i586-pc-cygwin-
+CROSS_TARGET=i686-pc-cygwin-
else
$(error Other Windows architectures not supported yet!)
endif
diff -r -u colinux-20040225.k6/src/colinux/os/winnt/build/Makefile colinux-20040225/src/colinux/os/winnt/build/Makefile
--- colinux-20040225.k6/src/colinux/os/winnt/build/Makefile 2004-02-26 03:29:09.109092800 +0600
+++ colinux-20040225/src/colinux/os/winnt/build/Makefile 2004-02-25 05:21:30.000000000 +0600
@@ -24,7 +24,7 @@
CLEAN_FILES += $(BUILD_PATH)/driver.base.tmp
$(BUILD_PATH)/driver.base.tmp: $(BUILD_PATH)/driver.o
- i586-pc-cygwin-gcc -Wl,--base-file,$@ \
+ i686-pc-cygwin-gcc -Wl,--base-file,$@ \
-Wl,--entry,_DriverEntry@8 \
-nostartfiles -nostdlib \
-o junk.tmp $^ -lntoskrnl -lhal -lgcc
@@ -32,12 +32,12 @@
CLEAN_FILES += $(BUILD_PATH)/driver.base.exp
$(BUILD_PATH)/driver.base.exp: $(BUILD_PATH)/driver.base.tmp
- i586-pc-cygwin-dlltool --dllname linux.sys \
+ i686-pc-cygwin-dlltool --dllname linux.sys \
--base-file $^ \
--output-exp $@
$(DRIVER_TARGET): $(BUILD_PATH)/driver.o $(BUILD_PATH)/driver.base.exp
- i586-pc-cygwin-gcc \
+ i686-pc-cygwin-gcc \
-Wl,--subsystem,native \
-Wl,--image-base,0x10000 \
-Wl,--file-alignment,0x1000 \
--
p.s.sorry for bad English
With regads,
Dmitriy mailto:dm...@ka...
|
|
From: Dan A. <da...@gm...> - 2004-02-26 10:12:32
|
On Wed, Feb 25, 2004 at 08:16:26PM -0500, Evan Danaher wrote: > I just wrote up a basic introduction to installing Colinux, which > hopefully is somewhat coherent and useful. It's nowhere near done, and > has not been proofread at all, but it's a start. For the moment, it's > just a static HTML file, but I'd be fine with making it CVS/wiki/whatever > so other people can work on it too. It's also at the moment basically > what I went through, and so is probably missing quite a bit of useful > information that might affect other people. But it's a start, and better > than nothing. > > http://www.tjhsst.edu/~edanaher/misc/colinux.html This is good, thanks for your effort. I really like Wiki as a platform for collaborated documentation, so I've setup a Wiki site on sourceforge: http://www.colinux.org/wiki/ Please add all your documentation to this site. -- Dan Aloni da...@gm... |
|
From: Nir P. <ni...@em...> - 2004-02-26 09:26:46
|
Hello, A question was asked on the IRC - can colinux support hardware devices directly ? I saw that linux/include/asm-i386/io.h is patched so outb/inb would not work, so I guess the answer is no. The real question is can the windows driver provide some generic i/o port service for the kernel, or can we somehow "share" the i/o space between windows ? Sorry in advance - I don't know Windows internals well enough to answer this myself. Nir |
|
From: Alex <ro...@ya...> - 2004-02-26 03:24:32
|
Hi,
I'm trying to run colinux 0.5.4 on dual boot system with debian on
reiserfs on /dev/hda2 and i'm getting these error messages during boot.
Similar messages i saw here recently, someone posted have them on ext3.
reiserfs: found format "3.6" with standard journal
Filesystem on 75:00 cannot be mounted because it is bigger than the device
You may need to run fsck or increase size of your LVM partition
Or may be you forgot to reboot after fdisk when it told you to
Kernel panic: VFS: Unable to mount root fs on 75:00
daemon: monitor terminated, reason 3
Can it be fixed somehow or maybe i'm doing something wrong ? Please,
advice.
My default.colinux.xml:
<?xml version="1.0" encoding="UTF-8"?>
<colinux>
<block_device index="0" path="\Device\HarddiskVolume2"
enabled="true"></block_device>
<bootparams>root=/dev/cobd0</bootparams>
<image path="vmlinux"></image>
<memory size="128"></memory>
<network index="0" type="tap" name="TAP"></network>
</colinux>
I builded vmlinux on debian unstable with gcc 3.3.2 and just reiserfs
turned on in .config compared to default shipped with colinux.
--
With best regards, Alex
|
|
From: Evan D. <eda...@tj...> - 2004-02-26 01:26:06
|
I just wrote up a basic introduction to installing Colinux, which hopefully is somewhat coherent and useful. It's nowhere near done, and has not been proofread at all, but it's a start. For the moment, it's just a static HTML file, but I'd be fine with making it CVS/wiki/whatever so other people can work on it too. It's also at the moment basically what I went through, and so is probably missing quite a bit of useful information that might affect other people. But it's a start, and better than nothing. http://www.tjhsst.edu/~edanaher/misc/colinux.html |
|
From: Alejandro R. S. <as...@MI...> - 2004-02-26 01:22:51
|
Yes, that's right. You can do just that if you're using the TAP network interface. On Wed, 2004-02-25 at 19:57, ch...@to... wrote: > I just mean that you can set windows to share the hardware > connection and windows will NAT to the tap connection. if the windows > machine is on a staticly assigned corporate lan that uses public addresses > and I am not the admin I can have windows assign a private ip to the > coLinux virtual network and provide NAT to access internet or real LAN. > > I guess the best thing would be for me to just give it a try when I get > some other things working the way I want. > > chris |
|
From: <ch...@to...> - 2004-02-26 00:50:46
|
so nothing to prevent connection sharing with tap right? I havn't messed with networking beyond comunicating with the host os yet. chris > > The bridged connection essentially 'shares' the host's NIC, using > winPCap to send out packets and listen for packets that it should > receive. Your dhcp example is right on, it would use the real dhcp > server. > > -Alejandro > > > > ------------------------------------------------------- > SF.Net is sponsored by: Speed Start Your Linux Apps Now. > Build and deploy apps & Web services for Linux with > a free DVD software kit from IBM. Click Now! > http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click > _______________________________________________ > coLinux-devel mailing list > coL...@li... > https://lists.sourceforge.net/lists/listinfo/colinux-devel > |
81 messages has been excluded from this view by a project administrator.
