platform/upstream/syslinux.git
13 years agocom32: Put deleted object files back into the Makefile
Matt Fleming [Wed, 13 Apr 2011 11:54:49 +0000 (12:54 +0100)]
com32: Put deleted object files back into the Makefile

For some reason that I can't understand, some object files were
removed from this Makefile in the following commit,

    |commit 0cb6490fa96f752c01bbb9add2c9ca6dbf99ece2
    |Author: Feng Tang <feng.tang@intel.com>
    |Date:   Thu Jun 3 10:48:22 2010 +0800
    |
    |   elflink: start merge with pathbased branch
    |
    |   modify these files to make compile pass
    |
    |       modified:   com32/MCONFIG
    |       modified:   com32/Makefile
    |       modified:   com32/include/klibc/compiler.h
    |       modified:   com32/include/sys/elfcommon.h
    |       modified:   com32/lib/Makefile
    |       modified:   com32/lib/free.c
    |       modified:   com32/lib/malloc.c

However, with this patch things continue to compile correctly so I'm
unsure why they were ever removed. Furthermore, pieces of code in
com32/ refer to symbols defined in these these object files and so
they are required in order to load modules.

These missing object files were discovered when loading libcom32.c32,
which wouldn't load because it was complaining about unresolved
symbols.

The removal of core/strncasecmp.c needs some explanation. There are
currently two identical copies of strncasecmp.c, one in com32/lib and
one in core/, and while the core copy is part of core/libcom32.a, the
copy in com32/lib isn't compiled. Now, because there are no references
to strncasecmp within core/ the symbol isn't pulled in from libcom32.a
and exported as a global symbol by the core. So, if ELF modules have
references to strncasecmp they cannot be resolved at runtime.

To fix this I've included strncasecmp.o in libcom32min.a which
isolinux.elf, ldlinux.elf and pxelinux.elf link against with
--whole-archive, which means that even if there are no references to
strncasecmp within the core, that symbol is still exported.

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agoelflink: Move more code from core/ into ldlinux
Matt Fleming [Fri, 8 Apr 2011 12:21:51 +0000 (13:21 +0100)]
elflink: Move more code from core/ into ldlinux

ldlinux now contains all the code necessary to load and execute
modules, none is contained in the core.

This change also allows us to change the spawn_load() prototype and to
push the job of processing arguments to executable functions (e.g. the
contents of char *argv[] as passed to a module's main function) into
ldlinux/execute.c instead of doing it in spawn_load(). Moving it into
ldlinux/ makes sense because the only core user of spawn_load() is
load_env32() and we don't require any sort of argument processing in
that path.

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agoelflink: Remove MODULE_MAIN() from executable ELF modules
Matt Fleming [Thu, 7 Apr 2011 20:02:02 +0000 (21:02 +0100)]
elflink: Remove MODULE_MAIN() from executable ELF modules

To make it easier to move existing COM32 modules over to ELF modules
without having to modify them we need to search for and run their
main() functions when they're loaded instead of requiring their
initial function to be labeled with MODULE_MAIN().

Currently, we require all executable ELF modules to specify their
initial function with the use of the MODULE_MAIN() macro and not via
the traditional method of naming it main(). However, there are weird
restrictions on what functions can be passed to MODULE_MAIN(), for
instance, they must be declared static.

This patch makes life much simpler and allows executable ELF modules
to be loaded without the MODULE_MAIN() wrapper, but rather by naming
their initial function main(). All the modules in com32/modules can
now be run as ELF modules without any modifications.

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agocom32: console is already initialised when we enter main()
Matt Fleming [Fri, 8 Apr 2011 12:52:30 +0000 (13:52 +0100)]
com32: console is already initialised when we enter main()

There's no need to initialise the console when we enter main() as
ldlinux.c32 will already have initialised it for us. Leaving the calls
to openconsole() causes the console settings to be reset whenever we
run a module.

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agoexit: Delete exit.S and references to __exit_handler
Matt Fleming [Tue, 12 Apr 2011 21:24:36 +0000 (22:24 +0100)]
exit: Delete exit.S and references to __exit_handler

We stopped using exit.S (and thus __exit_handler) in 71d5d1ba646e
("elflink: change the exit.c, remove the exit.S to make process exit
working") when we moved to using the C version of _exit(). If we don't
remove the symbol __exit_handler we'll encounter undefined symbol
errors when moving libcom32.a over to an ELF shared library.

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agomk: LDFLAGS use elf_i386 emulation for ld
Matt Fleming [Tue, 12 Apr 2011 21:23:55 +0000 (22:23 +0100)]
mk: LDFLAGS use elf_i386 emulation for ld

elf32_i386 isn't a valid ld emulation mode. Presumably no one is using
$(LDFLAGS) set in this file.

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agomk: Add rule for emitting .c32 ELF files
Matt Fleming [Thu, 7 Apr 2011 20:00:31 +0000 (21:00 +0100)]
mk: Add rule for emitting .c32 ELF files

We currently open code the Makefile rules for outputting .c32 files,
so provide a pattern rule to make things easier.

There are no users of this rule yet, but we'll make use of it when we
switch all modules over to ELF.

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agoelflink: Move elf MCONFIG to mk/elf.mk
Matt Fleming [Thu, 7 Apr 2011 17:12:51 +0000 (18:12 +0100)]
elflink: Move elf MCONFIG to mk/elf.mk

Catch up with recent changes to the Makefile fragments in the MCONFIG
files.

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agoMerge branch 'for-hpa/elflink/core' into elflink
Matt Fleming [Tue, 26 Apr 2011 08:51:32 +0000 (09:51 +0100)]
Merge branch 'for-hpa/elflink/core' into elflink

13 years agoMerge branch 'for-hpa/elflink/ldlinux' into elflink
Matt Fleming [Tue, 26 Apr 2011 08:51:14 +0000 (09:51 +0100)]
Merge branch 'for-hpa/elflink/ldlinux' into elflink

13 years agoMerge remote-tracking branch 'zytor/master' into elflink
Matt Fleming [Tue, 26 Apr 2011 08:46:34 +0000 (09:46 +0100)]
Merge remote-tracking branch 'zytor/master' into elflink

We need the recent Makefile filename changes to be merged into the
elflink branch because it will make things simpler when converting all
modules to ELF format.

Conflicts:
com32/Makefile
com32/modules/Makefile
version

13 years agohdt: add back -I$(com32)
H. Peter Anvin [Tue, 26 Apr 2011 08:30:29 +0000 (01:30 -0700)]
hdt: add back -I$(com32)

hdt picks up include files from outside the include directory and so
needs an explicit -I$(com32).  This probably should be fixed.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agosysdump: add $(LIBS) to the link rule
H. Peter Anvin [Tue, 26 Apr 2011 08:29:28 +0000 (01:29 -0700)]
sysdump: add $(LIBS) to the link rule

$(C_LIBS) doesn't include $(LIBS) so we need to add it explicitly.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agoMerge remote-tracking branch 'mfleming/for-hpa/makefile-cleanup'
H. Peter Anvin [Tue, 26 Apr 2011 08:18:18 +0000 (01:18 -0700)]
Merge remote-tracking branch 'mfleming/for-hpa/makefile-cleanup'

Resolved Conflicts:
com32/hdt/Makefile
com32/sysdump/Makefile

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agoisolinux: remove broken and obsolete open_file_t
H. Peter Anvin [Tue, 26 Apr 2011 03:18:07 +0000 (20:18 -0700)]
isolinux: remove broken and obsolete open_file_t

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agoisolinux: remove .img file support
H. Peter Anvin [Tue, 26 Apr 2011 03:16:07 +0000 (20:16 -0700)]
isolinux: remove .img file support

Remove disk image support from ISOLINUX.  Very few BIOSes support this
call correctly, and it is likely to have been broken for a very long
time anyway.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agocom32: inet_ntoa() takes struct in_addr
H. Peter Anvin [Tue, 26 Apr 2011 00:15:40 +0000 (17:15 -0700)]
com32: inet_ntoa() takes struct in_addr

The standard definition for inet_ntoa() is to take struct in_addr, and
not doing that causes a conflict on the lwip branch.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agohdt: remove set-but-unused variable
H. Peter Anvin [Tue, 26 Apr 2011 00:00:22 +0000 (17:00 -0700)]
hdt: remove set-but-unused variable

Remove a set but unused variable in HDT, which prevents building with
gcc 4.6.  This is probably a bug; presumably the active flag should be
presented to the user, too.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agoMerge remote-tracking branch 'sherbszt/gfxboot32'
H. Peter Anvin [Mon, 25 Apr 2011 22:24:33 +0000 (15:24 -0700)]
Merge remote-tracking branch 'sherbszt/gfxboot32'

13 years agogfxboot: pass current working dir to gfxboot core
Steffen Winterfeldt [Tue, 19 Apr 2011 13:27:26 +0000 (15:27 +0200)]
gfxboot: pass current working dir to gfxboot core

Pass current working dir to (and back from) gfxboot core via gfx_config
struct. This way I can implement a chdir() in the gfxboot core without
the need for additional 16-bit comboot calls. gfxboot may change the
directory, so you have to read and apply the current setting in boot() (to
read kernel etc. from right location).

Signed-off-by: Sebastian Herbszt <herbszt@gmx.de>
13 years agoMerge remote-tracking branch 'hdt/master'
H. Peter Anvin [Mon, 25 Apr 2011 21:34:30 +0000 (14:34 -0700)]
Merge remote-tracking branch 'hdt/master'

13 years agodiag/geodsp: zero-fill sectors in -1 mode
H. Peter Anvin [Mon, 25 Apr 2011 21:32:08 +0000 (14:32 -0700)]
diag/geodsp: zero-fill sectors in -1 mode

In -1 mode, zero-fill the sector instead of writing whatever garbage
was in the buffer already.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agodiag/geodsp: fix ordering of args to fread/fwrite in mk-lba-img
H. Peter Anvin [Mon, 25 Apr 2011 21:27:28 +0000 (14:27 -0700)]
diag/geodsp: fix ordering of args to fread/fwrite in mk-lba-img

If we want the return value to be in bytes, make sure we pass the byte
count in the nmemb field instead of the size field.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agodiag/geodsp: Avoid large intermediate files
H. Peter Anvin [Mon, 25 Apr 2011 19:39:08 +0000 (12:39 -0700)]
diag/geodsp: Avoid large intermediate files

Allow the generation program mk-lba-img to merge the prefix file and
output to stdout, so it can be fed directly to the compressor; that
way we avoid generating huge intermediate files on disk.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agoppmtolss16: remove obsolete Perl construct
H. Peter Anvin [Mon, 25 Apr 2011 19:22:21 +0000 (12:22 -0700)]
ppmtolss16: remove obsolete Perl construct

defined() on an array is considered obsolete; instead check for a
nonzero number of elements with scalar().

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agogfxboot: handle INITRD config line
Steffen Winterfeldt [Tue, 19 Apr 2011 13:27:26 +0000 (15:27 +0200)]
gfxboot: handle INITRD config line

Handle case where there is no initrd= option but a separate initrd config line.

Signed-off-by: Sebastian Herbszt <herbszt@gmx.de>
13 years agoversion: now working on version 4.05
H. Peter Anvin [Fri, 22 Apr 2011 04:32:01 +0000 (21:32 -0700)]
version: now working on version 4.05

4.04 is out, next version will be 4.05 (or later).

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agohdt: Final release 0.5.0
Erwan Velu [Thu, 21 Apr 2011 19:47:57 +0000 (21:47 +0200)]
hdt: Final release 0.5.0

Thanks for all who participe in this release.

13 years agohdt: Using dashed instead of . into label name
Erwan Velu [Thu, 21 Apr 2011 19:43:54 +0000 (21:43 +0200)]
hdt: Using dashed instead of . into label name

When generating floppy images some version of mtools (< 4.0.14) are
buggy and generate incorrect labels. So using > 4.0.13 is recommended.

Anyway, let's use _ into the volume name, that cleaner :

We now have
  disk label="HDT_0_5_0_B"
instead of
  disk label="HDT_0~5~0-B"

13 years agohdt: Fixing ACPI display
Erwan Velu [Wed, 20 Apr 2011 19:06:14 +0000 (21:06 +0200)]
hdt: Fixing ACPI display

Some screenshots shows that we have some troubles for aligning items.
This is now fixed.

13 years agohdt: New release beta5
Erwan Velu [Wed, 20 Apr 2011 18:44:29 +0000 (20:44 +0200)]
hdt: New release beta5

13 years agolibupload: Adding tftp.h missing file
Erwan Velu [Wed, 20 Apr 2011 18:21:21 +0000 (20:21 +0200)]
libupload: Adding tftp.h missing file

Oups, this while wasn't commited :/

13 years agohdt: Showing specific message for auto mode
Erwan Velu [Wed, 20 Apr 2011 18:20:22 +0000 (20:20 +0200)]
hdt: Showing specific message for auto mode

That's a nice info to know that some output were generated by the
automatic mode.

13 years agohdt: Removing useless printf in dump mode
Erwan Velu [Wed, 20 Apr 2011 18:14:45 +0000 (20:14 +0200)]
hdt: Removing useless printf in dump mode

This message was clearly boring to read ;o)

13 years agolibupload: Fixing tftp return value
Erwan Velu [Wed, 20 Apr 2011 18:12:24 +0000 (20:12 +0200)]
libupload: Fixing tftp return value

The default return value for the tftp mode.
This made the tftp to return error even if the upload was succesful.

13 years agohdt: New release Beta4
Erwan Velu [Tue, 19 Apr 2011 18:23:20 +0000 (20:23 +0200)]
hdt: New release Beta4

13 years agohdt: Adding dump mode on pre-built images
Erwan Velu [Tue, 19 Apr 2011 18:05:35 +0000 (20:05 +0200)]
hdt: Adding dump mode on pre-built images

As we now have a dump mode, let's add an entry in our menu for that.

13 years agohdt: Preventing dump mode if no PXE config
Erwan Velu [Tue, 19 Apr 2011 18:00:46 +0000 (20:00 +0200)]
hdt: Preventing dump mode if no PXE config

Dump mode can only work if PXE is present.

13 years agoMerge commit 'syslinux-4.04' into dynamic-sector
H. Peter Anvin [Mon, 18 Apr 2011 21:53:45 +0000 (14:53 -0700)]
Merge commit 'syslinux-4.04' into dynamic-sector

13 years agospec: Use BuildRequires, add diag/ subdirectory syslinux-4.04
H. Peter Anvin [Mon, 18 Apr 2011 21:24:17 +0000 (14:24 -0700)]
spec: Use BuildRequires, add diag/ subdirectory

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agodiag: Don't remove BTARGET on make clean
H. Peter Anvin [Mon, 18 Apr 2011 21:12:58 +0000 (14:12 -0700)]
diag: Don't remove BTARGET on make clean

"make clean" should not remove BTARGET

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agoMerge branch 'master' of git://git.kernel.org/pub/scm/boot/syslinux/syslinux
Erwan Velu [Mon, 18 Apr 2011 21:12:22 +0000 (23:12 +0200)]
Merge branch 'master' of git://git./boot/syslinux/syslinux

13 years agoMerge remote-tracking branch 'pcacjr/syslinux-update-for-hpa'
H. Peter Anvin [Mon, 18 Apr 2011 21:11:26 +0000 (14:11 -0700)]
Merge remote-tracking branch 'pcacjr/syslinux-update-for-hpa'

13 years agohdt: Reporting TFTP errors in dump mode
Erwan Velu [Mon, 18 Apr 2011 21:08:21 +0000 (23:08 +0200)]
hdt: Reporting TFTP errors in dump mode

If we get some errors from the tftp, let's report it to the user.
This will make user's life easier to understand why the dumping mode
isn't working.

13 years agolibupload: Implementing error code on tftp
Erwan Velu [Mon, 18 Apr 2011 21:07:39 +0000 (23:07 +0200)]
libupload: Implementing error code on tftp

When using the tftp backend, let's return the errors we got on the
server.

13 years agomenu: allow "menu hiddenkey" to take multiple keys; run unlabel
H. Peter Anvin [Mon, 18 Apr 2011 21:06:54 +0000 (14:06 -0700)]
menu: allow "menu hiddenkey" to take multiple keys; run unlabel

- Allow "menu hiddenkey" to take a comma-separated list of keys.
- Run unlabel() on the commands passed to menu hiddenkey.
  XXX: Consider moving unlabelling to post-menu instead.

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agohdt: Fixing automatic mode parsing
Erwan Velu [Mon, 18 Apr 2011 20:37:15 +0000 (22:37 +0200)]
hdt: Fixing automatic mode parsing

Parsing of the auto='' mode was weak an non functionnal in many
configuration like

auto='dump'

This is now fixed.

13 years agoNEWS: Document MENU HIDDENKEY
H. Peter Anvin [Mon, 18 Apr 2011 20:53:19 +0000 (13:53 -0700)]
NEWS: Document MENU HIDDENKEY

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agolibupload: Fixing memset call
Erwan Velu [Mon, 18 Apr 2011 19:07:33 +0000 (21:07 +0200)]
libupload: Fixing memset call

We have to use the size of the element, not the size of the pointer.

13 years agohdt: New release 0.5.0-beta3
Erwan Velu [Sun, 17 Apr 2011 17:44:21 +0000 (19:44 +0200)]
hdt: New release 0.5.0-beta3

13 years agocpuid: Using generic model if cyrix id is unknown
Erwan Velu [Sat, 16 Apr 2011 18:34:40 +0000 (20:34 +0200)]
cpuid: Using generic model if cyrix id is unknown

When using the specific Cyrix call to determine the model, if the model
is unknown, let's consider the generic call was right.

This insure a better display of the model id.

13 years agoldlinux: Pass entire cmdline to execute()
Matt Fleming [Fri, 8 Apr 2011 12:15:43 +0000 (13:15 +0100)]
ldlinux: Pass entire cmdline to execute()

We need to pass the entire cmdline to execute(), not just the kernel
name. execute() does all the required processing of cmdline arguments
and passes them to the loader in the correct format, e.g. for COM32
modules the cmdline arguments need to be passed in the traditional
char *argv[] way.

So stop using strtok() which modifies its first argument and means we
therefore lose the cmdline after the kernel name.

Previously, the "argc" and "argv" arguments that are passed to a
module's main() function contained bogus values.

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agoldlinux: Move cmdline processing out of cli.c
Matt Fleming [Wed, 6 Apr 2011 13:37:43 +0000 (14:37 +0100)]
ldlinux: Move cmdline processing out of cli.c

While it makes perfect sense for cli.c to read input from the user and
parse certain characters (Ctrl-R, TABs, etc), once that processing is
done it should be up to the caller to decide what to do with the
expanded cmdline string. For example, the logic to figure out what
type of kernel we're loading (calculated from a kernel's filename
extension as typed on the command-line) is currently in cli.c, and
that is just wrong. That kind of logic belongs in ldlinux.c

So move process_command() from cli.c to ldlinux.c and extend it so
that if the user types a label name at the prompt, we load the kernel
for that label as described in the config file.

Now that we've moved all cli history code into edit_cmdline() it seems
like a good time to turn it on permanently.

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agoldlinux, cli: Do not assign 'comm_counter' unnecessarily
Matt Fleming [Wed, 6 Apr 2011 11:59:29 +0000 (12:59 +0100)]
ldlinux, cli: Do not assign 'comm_counter' unnecessarily

There's no need to assign a value to 'comm_counter' at the start of
the edit_cmdline() because it is always assigned before being used.

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agoelflink: Jump to kaboom if we fail to load ldlinux.c32
Matt Fleming [Wed, 6 Apr 2011 12:31:16 +0000 (13:31 +0100)]
elflink: Jump to kaboom if we fail to load ldlinux.c32

Previously if we failed to load ldlinux.c32, we would drop out of
load_env32() and try to execute the assembly kernel loader.

With this patch, if we drop out of load_env32() we jump to kaboom
which drops the user at a command-line, at least that way they have a
chance of fixing the reason ldlinux.c32 failed to load.

Also, delete the comment that talks about loading and parsing the
configuration file - that is now done by ldlinux.c32.

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agolua: Delete superfluous $(LIBS) prerequisite
Matt Fleming [Tue, 12 Apr 2011 13:44:32 +0000 (14:44 +0100)]
lua: Delete superfluous $(LIBS) prerequisite

com32.mk already includes the necessary library in $(C_LIBS), we don't
need to list them again in $(LIBS).

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agosysdump: Delete superfluous $(LIBS) prerequisite
Matt Fleming [Tue, 12 Apr 2011 13:39:57 +0000 (14:39 +0100)]
sysdump: Delete superfluous $(LIBS) prerequisite

com32.mk already includes the necessary libraries in $(C_LIBS), we
don't need to list them again in $(LIBS).

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agohdt: Delete superfluous $(LIBS) prerequisites
Matt Fleming [Mon, 11 Apr 2011 13:44:44 +0000 (14:44 +0100)]
hdt: Delete superfluous $(LIBS) prerequisites

com32.mk already includes the necessary libraries in $(C_LIBS), we
don't need to list them again in $(LIBS).

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agomenu.c32: Delete superfluous $(LIBS) prerequistite
Matt Fleming [Mon, 11 Apr 2011 12:38:36 +0000 (13:38 +0100)]
menu.c32: Delete superfluous $(LIBS) prerequistite

com32.mk already includes the necessary libraries in $(C_LIBS), we
don't need to list them again in $(LIBS).

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agocpuid: Improving Cyrix/NSC detection
Erwan Velu [Sat, 16 Apr 2011 06:06:01 +0000 (08:06 +0200)]
cpuid: Improving Cyrix/NSC detection

This code add the specific detection code for Cyrix/NSC processor.
Code came from the Linux kernel.

13 years agohdt: Fixing buffer corruption in PXE menu
Erwan Velu [Fri, 15 Apr 2011 19:26:52 +0000 (21:26 +0200)]
hdt: Fixing buffer corruption in PXE menu

Jim Cromie reported that PXE menu behave like :

gPXE Detected: ]No

This commit avoid this buffer corruption.

13 years agosyslinux: check --update option properly
Paulo Alcantara [Fri, 15 Apr 2011 18:28:31 +0000 (15:28 -0300)]
syslinux: check --update option properly

Signed-off-by: Paulo Alcantara <pcacjr@gmail.com>
13 years agoextlinux: remove already_installed
Paulo Alcantara [Fri, 15 Apr 2011 18:21:10 +0000 (15:21 -0300)]
extlinux: remove already_installed

Signed-off-by: Paulo Alcantara <pcacjr@gmail.com>
13 years agoextlinux: use syslinux_already_installed instead of already_installed
Paulo Alcantara [Fri, 15 Apr 2011 18:13:54 +0000 (15:13 -0300)]
extlinux: use syslinux_already_installed instead of already_installed

Signed-off-by: Paulo Alcantara <pcacjr@gmail.com>
13 years agolibinstaller: implement syslinux_already_installed
Paulo Alcantara [Fri, 15 Apr 2011 17:55:29 +0000 (14:55 -0300)]
libinstaller: implement syslinux_already_installed

syslinux_already_installed function will be used in both extlinux and
syslinux for checking if the boot sector has either the string
"SYSLINUX" or "EXTLINUX" in the OEMID field.

Signed-off-by: Paulo Alcantara <pcacjr@gmail.com>
13 years agocpuid: Managing UNKNOWN CPU better
Erwan Velu [Thu, 14 Apr 2011 18:58:05 +0000 (20:58 +0200)]
cpuid: Managing UNKNOWN CPU better

It's safe to manage the unknown cpu as a regular one. This will avoid
weird behavior when CPU isn't well detected.

13 years agohdt: Changing boot flag semantic in menu mode
Erwan Velu [Wed, 13 Apr 2011 19:13:37 +0000 (21:13 +0200)]
hdt: Changing boot flag semantic in menu mode

What we report is the presence of the boot flag and not the ability of a
partition to be bootable. A partition can have a valid bootloader with a
boot flag not set. So the 'Bootable' semantic was wrong.

13 years agocpuid: add cpuid info for "Geode by NSC"
Jim Cromie [Wed, 13 Apr 2011 19:03:58 +0000 (21:03 +0200)]
cpuid: add cpuid info for "Geode by NSC"

Without this, cpu_detect(), and set_generic_info() would take 416
seconds on a soekris 4801 and hdt would report ?????????? as the
processor vendor.

13 years agohdt: fix precedence problem in double-space check
Jim Cromie [Wed, 13 Apr 2011 18:57:31 +0000 (20:57 +0200)]
hdt: fix precedence problem in double-space check

13 years agohdt: Display non-sequentials disks configurations
Erwan Velu [Wed, 13 Apr 2011 18:42:38 +0000 (20:42 +0200)]
hdt: Display non-sequentials disks configurations

If the location of the disks wasn't sequential on the system, the menu
mode was missing some.

This commit fix ticket #46.

Thanks gerth for reporting.

13 years agohdt: Adding chain.c32 to pre-built images
Erwan Velu [Wed, 13 Apr 2011 17:33:27 +0000 (19:33 +0200)]
hdt: Adding chain.c32 to pre-built images

This got a requirement from some users.

13 years agohdt: Fixing checksums deletion
Erwan Velu [Tue, 12 Apr 2011 17:43:34 +0000 (19:43 +0200)]
hdt: Fixing checksums deletion

We have to delete all the checksums files while cleaning.
Older release might exists.

13 years agohdt: New release 0.5.0-beta2
Erwan Velu [Tue, 12 Apr 2011 17:28:38 +0000 (19:28 +0200)]
hdt: New release 0.5.0-beta2

13 years agokeytest: make somewhat more comprehensible
H. Peter Anvin [Mon, 11 Apr 2011 22:42:06 +0000 (15:42 -0700)]
keytest: make somewhat more comprehensible

Someone may very well need to use keytest to figure out their
keymappings.  Make it a bit more user friendly.

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agokeyname: better handling of named control characters
H. Peter Anvin [Mon, 11 Apr 2011 22:41:26 +0000 (15:41 -0700)]
keyname: better handling of named control characters

Named control character (Tab, Enter, Backspace etc.) should use those
names rather than caret sequences.

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agokeytest: show the human name together with the keycode
H. Peter Anvin [Mon, 11 Apr 2011 22:33:48 +0000 (15:33 -0700)]
keytest: show the human name together with the keycode

Since we now have library support for showing the keycode name, make
keytest print that as well.

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agomenu: clear the timeout line on MENU HIDDENKEY
H. Peter Anvin [Mon, 11 Apr 2011 22:23:05 +0000 (15:23 -0700)]
menu: clear the timeout line on MENU HIDDENKEY

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agomenu hiddenkey: fix parsing the config file option
H. Peter Anvin [Mon, 11 Apr 2011 22:22:17 +0000 (15:22 -0700)]
menu hiddenkey: fix parsing the config file option

Fix parsing bugs in the config file.

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agomenu: Add "menu hiddenkey" to make terminating menu hidden a command
H. Peter Anvin [Mon, 11 Apr 2011 21:59:45 +0000 (14:59 -0700)]
menu: Add "menu hiddenkey" to make terminating menu hidden a command

Use "menu hiddenkey" together with "menu hidden" for a one-keystroke
action.

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agohdt: Adding dump entry in the menu
Erwan Velu [Mon, 11 Apr 2011 20:02:52 +0000 (22:02 +0200)]
hdt: Adding dump entry in the menu

If the PXE is enabled, let's show a dump menu at the main menu.

13 years agohdt: Adding checksums file
Erwan Velu [Mon, 11 Apr 2011 19:29:27 +0000 (21:29 +0200)]
hdt: Adding checksums file

When generating a release, let's create a list of the checksums

13 years agohdt: Adding dump_mode & tftp_ip boot option
Erwan Velu [Mon, 11 Apr 2011 19:10:11 +0000 (21:10 +0200)]
hdt: Adding dump_mode & tftp_ip boot option

dump_mode=<dir> give the user a chance to select a directory on the tftp
server.

tftp_ip=<w.x.y.z> give the user a chance to select another tftp server
for dumping data. By default, we use the tftp that serves the pxe
booting.

13 years agohdt: new beta 0.5.0-beta1
Erwan Velu [Mon, 11 Apr 2011 18:32:49 +0000 (20:32 +0200)]
hdt: new beta 0.5.0-beta1

This first beta is featuring the dumping code.

13 years agohdt: Dumping HDT
Erwan Velu [Mon, 11 Apr 2011 18:32:27 +0000 (20:32 +0200)]
hdt: Dumping HDT

HDT information is now dumped in the 'hdt' file

13 years agohdt: Adding Kernel dumping
Erwan Velu [Mon, 11 Apr 2011 18:09:52 +0000 (20:09 +0200)]
hdt: Adding Kernel dumping

Kernel is now dumped into th 'kernel' file

13 years agoMerge remote branch 'hpa/master' into libupload
Erwan Velu [Mon, 11 Apr 2011 17:41:56 +0000 (19:41 +0200)]
Merge remote branch 'hpa/master' into libupload

13 years agomboot.c32: Delete superfluous $(LIBS) prerequisite
Matt Fleming [Mon, 11 Apr 2011 12:17:32 +0000 (13:17 +0100)]
mboot.c32: Delete superfluous $(LIBS) prerequisite

com32.mk already includes the necessary archive libraries in
$(C_LIBS), we don't need to list them again in $(LIBS).

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agoMakefile: Move Makefile fragments into mk/
Matt Fleming [Thu, 7 Apr 2011 16:12:08 +0000 (17:12 +0100)]
Makefile: Move Makefile fragments into mk/

Move the MCONFIG files into a mk/ directory and give them more
descriptive names.

This is purely a cosmetic change to make the 'include' directives a
bit more coherent by making it obvious exactly which MCONFIG file
we're including. For example, in com32/lua/src/Makefile we exchange
the line,

include ../../MCONFIG

for the much more comprehensible,

include $(MAKEDIR)/com32.mk

Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
13 years agodiskstart: use the correct register for the starting address
H. Peter Anvin [Tue, 5 Apr 2011 21:35:55 +0000 (14:35 -0700)]
diskstart: use the correct register for the starting address

The starting address is obtained in si, not edi.

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agoinstaller: Update and correct generate_extents()
H. Peter Anvin [Tue, 5 Apr 2011 21:24:07 +0000 (14:24 -0700)]
installer: Update and correct generate_extents()

1. Make sure extents are strictly less than 64K in size
2. Make sure extents do not cross 64K boundaries

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agocore: Move sector 1 to address 0x8000 so it is aligned
H. Peter Anvin [Tue, 5 Apr 2011 21:07:32 +0000 (14:07 -0700)]
core: Move sector 1 to address 0x8000 so it is aligned

Move sector 1 to 0x8000, so that it will be sector-aligned no matter
what the sector size is (well, as long as it is <= 32K).  This avoids
potential conflicts between the 64K "seams" and the sector size.

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agocore/head.inc: require NASM 2.03+
H. Peter Anvin [Tue, 5 Apr 2011 21:07:06 +0000 (14:07 -0700)]
core/head.inc: require NASM 2.03+

We have needed NASM 2.03+ for a long time.

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agoMerge commit 'syslinux-4.04-pre22' into dynamic-sector
H. Peter Anvin [Tue, 5 Apr 2011 20:58:30 +0000 (13:58 -0700)]
Merge commit 'syslinux-4.04-pre22' into dynamic-sector

13 years agodiskstart: Add a pointer to the syslinux_banner to the EPA syslinux-4.04-pre22
H. Peter Anvin [Tue, 5 Apr 2011 00:38:18 +0000 (17:38 -0700)]
diskstart: Add a pointer to the syslinux_banner to the EPA

For the benefit of third-party tools, add a pointer to the
syslinux_banner in the extended patch area.

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agodisk: Make the sector size dynamic
H. Peter Anvin [Tue, 5 Apr 2011 00:26:59 +0000 (17:26 -0700)]
disk: Make the sector size dynamic

Make the sector size dynamic in anticipation of disks with
non-512-byte logical sectors.

With this change, bsBytesPerSec *must* be correct; this still needs to
be added to the installers to enforce.  Furthermore, the number of
sectors in one extent can still only be < 64K.

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agocom32, bitops: proper constraint for byte-sized output syslinux-4.04-pre21
H. Peter Anvin [Mon, 4 Apr 2011 23:43:10 +0000 (16:43 -0700)]
com32, bitops: proper constraint for byte-sized output

The proper output for the byte-sized output of "setc" is "=qm".

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agoNEWS: iso9660 also had issues with ../ paths. syslinux-4.04-pre20
H. Peter Anvin [Mon, 4 Apr 2011 23:31:55 +0000 (16:31 -0700)]
NEWS: iso9660 also had issues with ../ paths.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
13 years agocore: Adjust formatting around the date string
H. Peter Anvin [Mon, 4 Apr 2011 23:22:38 +0000 (16:22 -0700)]
core: Adjust formatting around the date string

Adjust the formatting around the date string to avoid missing and
duplicated spaces.

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agotinyjpeg: change DEBUG to JPEG_DEBUG
H. Peter Anvin [Mon, 4 Apr 2011 23:17:48 +0000 (16:17 -0700)]
tinyjpeg: change DEBUG to JPEG_DEBUG

Rename the debugging option for tinyjpeg to JPEG_DEBUG to avoid
conflict with the global dprintf option.

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
13 years agoiso9660: use generic_mangle_name
H. Peter Anvin [Mon, 4 Apr 2011 23:14:37 +0000 (16:14 -0700)]
iso9660: use generic_mangle_name

There is no longer any need to do special name-mangling for iso9660,
and doing so, in fact, breaks the handling of paths starting with ../.

Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>