freesource
0eadd6eec2
Tested gbootroot-nest1 within a uml, nice, will have to tweak the
logic for cmdline only to set-up .gbootroot* properly, and the logic
pertaining to mode={tt,skas}, but other than that things are working
nicely. Will have to come out with a new make-debian-x11 package, but
then again will need to come out with a new root_fs representing the
nesting changes, too.
22 years ago
freesource
28ba4c9b94
Preping for release.
22 years ago
freesource
405d19a98e
This gets things working nicely. Now exits sooner for root_fs_helper
because the nbd line no longer presents itself. Makefile is almost
done.
22 years ago
freesource
7df8826e1e
Made linux, linuxbr to avoid overlapping with other packages, ofcource
this could be made a variable, too, in BootRoot.pm.
22 years ago
freesource
81051244ae
Just properly orders --help a little
22 years ago
freesource
953854a5a4
This gets everything working properly. Now --home works properly both
for gui-mode and cli, and all executables can be properly specified.
Ex:
./gbootroot --home . --template Example-Mini.yard --uml-exclusively on
--expect-program ./expect_uml
--uml-kerneluser-mode-linux/usr/bin/linux --root-fs-helper
root_filesystem/root_fs_helper
./gbootroot --home . --template Example-Mini.yard --genext2fs-dir
genext2fs/ --filesystem-command "genext2fs -z -r0"
22 years ago
freesource
29e9e73202
This allows adds the kernel option for UML in expect_uml, it's amazing
what you find when you remove the system package. Getting closer.
22 years ago
freesource
50bbbb674c
cvs from emacs
22 years ago
freesource
6ad8077da8
documented genext2fs-dir
22 years ago
freesource
4c90daea57
Changed linuxbr to linux, may ultimately change it to linux so it doesn't
conflict with user-mode-linux, need to check on the set-up. Probably gbootroot
should be made to conflict with these deb packages. Options has more
explanation and another hidden option added.
22 years ago
freesource
8827b1b0d4
Some even more elaborate changes so that the gui can be ran from the immediate
directory for testing using only the --home option. Also, made sure genext2fs
can be found.
22 years ago
freesource
705d8a7856
Now there are some hidden options so that root_fs can be made from the
programs from within the source code like:
perl -I . ./gbootroot --home . --template Helper.yard --root-filename root_fs_experiment --root-fs-helper-location ./root_filesystem/root_fs_helper --uml-exclusively on --expect-program ./expect_uml
22 years ago
freesource
bfed2bb88f
This corrects nss_check so that the proper version from libc6 is recognized
for nss libraries. Hopefully, this corrects problems for even newer libc6
versions in the future.
22 years ago
freesource
578ddd6dd2
Now things are working perfectly. Had to add a last statement to the while
wrapper for expect_uml, this is a new bug. And added preserve-permissions.
22 years ago
freesource
7727285976
Didn't do a save.
22 years ago
freesource
b78a85686f
This add --help|-h and the corresponding text. Just need to get uml-exlusively
working properly
22 years ago
freesource
407879252c
Changed stdout to no-stdout so that stdout is the default, but can be turned
off, and fixed a major bug with undef value used in dirname which caused to
program to die.
22 years ago
freesource
f960c66fcb
This gets just about every option working, except for uml-kernel and
preserve-ownership. Adds color --stdout. Basically, the bare min is
--template and --stdout for display on console.
22 years ago
freesource
2f4e1cdcc6
This gets read_contents_file working, hopefully error() is the only gtk related
sub to watch out for.
22 years ago
freesource
830a1c109a
Now gbootroot is functioning well as either a GUI or CLI in relation to whether or not arguments are added. Options.pm allows things to be checked before interpreting so that a determination can be made what mode to go in.
22 years ago
freesource
8db50ec6ac
This adds some more doc, and begins to explore GetOpt, unfortunately it's presently going into an endless loop when there is an incorrect opt.
22 years ago
freesource
79631e765c
Planning documentaion for variables for commandline.
22 years ago
freesource
9de07f7f5e
This is the start of giving gbootroot some commandline functionality similar to
bootroot, but more specific with the primary goal of allowing root_fs
prototyping without the use of the GUI.
22 years ago
freesource
d077def592
Improved the parameter explanation .. ramdisk_size really wasn't crucial, and
removed mem=16M as a default, first of all this is already standard, and secondly
it may cause problems for users unaware of the MTD Emulators default settings.
23 years ago
freesource
59c8a5333e
This only checks Root_Fs for $ubd0, which makes a lot more sense and avoides
unecessary duplications.
23 years ago
freesource
bcf2987f21
This adds better memory ratios for blkmtd.
23 years ago
freesource
63dc15fe8e
Improves the mtd_init regexp even more.
23 years ago
freesource
2e41744e08
This makes for a better regexp .. will also match "/sbin/init 2" for mtd_init
23 years ago
freesource
24897ac61e
This goes a step farther an allows ubd to be ubd\d{1] rather then confined to
1, however, mtd and mtdblock will be 0 for the first module invocation.
23 years ago
freesource
80ebd88aef
This allow ubd0\w?, this works with mtdram, but causes blkmtd to fail.
23 years ago
freesource
86f767a446
Yard won't attempt to make /dev on compressed fs and ro if /dev is a symbolic
link, won't complain when there is an unitialized value in concatation in
Boot, Ramdisk_size is always 4096 for mtdram, this setting isn't really
necessary, but is left anyways.
23 years ago
freesource
4da04f8623
This adds ceil() to $total_size, and always makes sure that ramdisk_size is
at least 4096.
23 years ago
freesource
bf4200e3c5
Fixed a little bug where initrd= wasn't being passed. Everythings looks good
with the default Initrd.gz.
23 years ago
freesource
ecc03b1980
This completes the init and mtd_init recognization .. added single.
23 years ago
freesource
967df903ea
This adds the mtd_init option to the kernel command-line so that the
pivot_root routine can be interacted with, and also checks for alternative
runlevels, all assigned to $init.
23 years ago
freesource
74dd88069a
This is even nicer. It gives both mtdram and blkmtd the same memory parameters
for mem. Just have to watch out for that nasty cramfsck by Linus.
23 years ago
freesource
fadcb89061
More mem adjustments, this time to the spin wheel, 1024 for 1, 8192 for 2,
these are great setting for mtdram .. <16M files can use the small adj.
>16M should zero down with 3 and increment with 2nd button.
23 years ago
freesource
30a11e4bf2
A few more improvements to the mtd code to handle memory better 16M increments
for mtdram, will probably do a similiar thing for blkmtd which appears to be
exponential 16 32 64 etc .. ofcourse at 128 things can get narly.
23 years ago
freesource
6b3eda4286
This pretty much covers it, have both mtdram and blkmtd working, just need to
figure out where Initrd.gz will be stored.
23 years ago
freesource
7fcb2ca973
Worked out lots of issues with mem, even using POSIX, tested well for mtdram,
interface is almost there.
23 years ago
freesource
07694a9bf1
This sets up everything, and tests the new mtd interface, looks good.
23 years ago
freesource
f7921951e1
This keeps a nice size ration for $mtd_total_size, and automatically checks if
the root_fs is real.
23 years ago
freesource
3fcea5beb5
Now it has the look.
23 years ago
freesource
4920a10f03
Better sizing.
23 years ago
freesource
9de4fc1d56
Remembers state for the new mtd emulation.
23 years ago
freesource
935b5983c8
Added a MTD emulation line to the UML box and some functionality.
23 years ago
freesource
bf98923042
This properly deals with differening mconsole locations.
23 years ago
freesource
896f2a9519
Added cad "Ctrl-Alt-Del" to mconsole interface, and documented it.
Now looks for mconsole at both old location /tmp/uml/$umid/mconsole and
new location $ENV{HOME}/.uml/$umid/mconsole.
23 years ago
freesource
93b5f23410
Now files for uml_exclusively are made in create_expect_uml which is
effectively known as the create stage. This avoids the unecessary previous
requirement of having to select the copy stage again just to create a file
for /dev/ubd/0. The file created is sparse, got this idea from a discussion
Jeff had with a user on uml-user.
23 years ago
freesource
d2fb7540f3
This adds jffs/jffs2 construction, expect_uml will have to be changed a little
bit more, but have tested this with an initrd, and it works pretty nicely
even with large fs assuming mem is set high enough. Mtdram was chosen over
blkmtd for now.
23 years ago