| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
| |
Clean up the handling of firmware selection inside the Makefiles.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Users are starting to package all Syslinux versions on one boot medium
and are running into problems because all versions of ldlinux have the
same filename, ldlinux.c32. The version of ldlinux.c32 compiled for bios
will not work with efi32 or efi64, and vice versa. What we need is a way
for the Syslinux core to know exactly which filename to load.
Use the following extensions for the corresponding platforms,
.c32 - bios
.e32 - efi32
.e64 - efi64
Now all ldlinux variants can be kept in the same directory without
causing any conflicts.
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
|
| |\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Conflicts:
Makefile
NEWS
com32/cmenu/Makefile
com32/elflink/ldlinux/Makefile
com32/gfxboot/Makefile
com32/gpllib/Makefile
com32/include/sys/module.h
com32/lib/Makefile
com32/lib/sys/module/elf_module.c
com32/menu/Makefile
com32/rosh/Makefile
com32/samples/Makefile
core/init.c
mk/elf.mk
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Tagging __syslinux_get_serial_info() with __constructor is pretty
useless when the global variables it uses, such as SerialPort, etc,
are assigned *after* the constructor has run. This constructor made
sense when config parsing was done by the core, but parsing is now
performed by ldlinux. We need to explicitly invoke the function to
initialise __syslinux_serial_console_info once we've parsed any config
files.
eprintf.c was introduced in commit 086d698c642f ("ldlinux: Add
eprintf() to print to VGA and serial") because printf() output wasn't
appearing on the serial console. It turns out that the above
__constructor confusion was the real bug.
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
|
| | |
| |
| |
| |
| |
| |
| |
| | |
Always strip the modules; they are too big unstripped. Specifically,
we generate unstripped *.elf files, and then convert them to
stripped *.c32 files.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
|
| |\ \
| |/
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Merge the changes since Syslinux-5.00. The main reason behind this
merge is to pull in the following bug fix, commit 90b5864fc793
("core/kaboom.c: Export __bad_SEG()"), which allows us to turn on
debugging in mk/devel.mk without ldlinux.c32 failing to load.
Conflicts:
com32/elflink/ldlinux/Makefile
version
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The documented command sequence for distributions wishing to package a
Syslinux release is,
make clean
make installer
Unfortunately, becaues ldlinux.c32 is deleted by 'make clean' the
installer target fails like so,
make[1]: *** No rule to make target `../com32/elflink/ldlinux/ldlinux.c32', needed by `ldlinuxc32_bin.c'. Stop.
make: *** [installer] Error 2
Follow the example set by core/Makefile for ldlinux.sys, etc and add
any build files that are required by the installers to
$(BTARGET). $(BTARGET) only gets deleted when running 'make spotless'.
Reported-by: László Házy <hazy_l@yahoo.com>
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
|
| | |
| |
| |
| |
| | |
Set the sonames of shared libraries to the filename, so we don't end
up with full pathnames embedded in the files.
|
| |\ \
| |/
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Conflicts:
Makefile
com32/elflink/ldlinux/Makefile
com32/lib/sys/module/elf_module.c
core/cleanup.c
core/comboot.inc
core/conio.c
core/fs/fs.c
core/init.c
core/mem/free.c
core/mem/malloc.c
core/timer.inc
diag/geodsp/Makefile
extlinux/main.c
mk/embedded.mk
modules/Makefile
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Before modules were dynamically loaded the boundary between GPL and
non-GPL code was implicit because of the separate link domains for
each module. With dynamic modules we need an explicit whitelist of
core symbols that non-GPL code can link against at runtime without
needing to be re-licensed under the GPL.
Mark such symbols with __export, so that it is explicitly clear which
symbols in the core can be linked against by non-GPL code.
Reduce the visibility of symbols in both the core and ldlinux.c32 with
-fvisibility=hidden. __export changes the visibility to 'default'.
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The code that handles the DISPLAY directive was writing directly to
the BIOS VGA page with __intcall(0x10). This caused corruption
problems on the screen because the ansi library code was also writing
to the screen.
The correct way to fix this is to always use the ansi library code
(via printf()) instead of going behind its back and using separate
code paths to write to the screen.
Reported-by: Ady <ady-sf@hotmail.com>
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
|
| |\ \
| |/
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Pull in the latest prerelease that includes some bug fixes for ldlinux
and a realloc() bug in core/.
Conflicts:
com32/cmenu/Makefile
com32/elflink/ldlinux/Makefile
com32/lib/Makefile
core/pxelinux.asm
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
commit 4503e6260c0 ("ldlinux: Use stdcon instead of rawcon for the
console") broke get_key() because it was no longer possible to read
raw data from stdin. Provide a new function raw_read() that will
return user input a character at a time.
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
|
| |/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Now that we have multiple firmware backends it no longer makes sense
to write object files to the same directory as their source. A better
solution is to write the object files to a per-firmware directory
under a top-level object directory.
The top-level object directory can be specified on the command-line
with the O= variable, e.g. make O=/tmp/obj. If no top-level object
directory is specified an 'obj' directory is created in the top-level
of the Syslinux source repository.
All the existing make targets continue to work as before, however now
they apply to all firmware backends, e.g. 'make installer' will build
the BIOS, 32-bit EFI and 64-bit EFI installers and place them under
$(OBJ)/bios, $(OBJ)/efi32 and $(OBJ)/efi64 respectively.
Note unlike every other bit of Syslinux, the gpxe objects are still
kept in the src directory, e.g. gpxe/src, since gpxe is only required
by the BIOS backend.
It is possible to specify a make target for a specific firmware or
list of firmware with the following syntax,
make [firmware[,firmware]] [target[,target]]
To clean the object directory for just the BIOS firmware type,
'make bios clean'
To build both the 32-bit and 64-bit EFI installers type,
'make efi32 efi64 installer'
Since the Syslinux make infrastructure is now more complex a new file
doc/building.txt has been created to explain how to build Syslinux.
The top-level Makefile now exports some make variables for use in
module Makefiles,
- topdir - the top-level source directory of the Syslinux
repository, e.g. /usr/src/syslinux
- objdir - the top-level object directory for the firmware
backend currently being built, e.g. /obj/syslinux/bios
- SRC - the source directory in the Syslinux repository for the
module currently being built,
e.g. /usr/src/syslinux/com32/libupload
- OBJ - the object directory for the module currently being
built, e.g. /obj/syslinux/bios/com32/libupload
Since we're rewriting the Makefile infrastructure anyway it seemed
like a good idea to add parallel support. By writing subdirectories as
prequisites for make targets the objects in those subdirectories can
be built in parallel.
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
|
| |
|
|
|
|
| |
To improve readability, move chainbooting to its own source file.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
This adds missing support for booting from a boot sector file such as
.bs, .bss or .0, by re-implementing the old asm bootsec code from
core/bootsect.inc in C.
This has resulted in some external changes. We've had to make StackBuf
a global symbol because we access it directly from execute.c. Also, we
need to move dsinfo.c into MINLIBOBJS because ldlinux now needs to
reference __syslinux_derivative_info.
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The old way of specifying functions that need to be run before/after a
main function has never worked for ELF modules. Instead, the only way
to get similiar behaviour was by using the MODULE_INIT() and
MODULE_EXIT() macros, but no one seems to have bothered converting the
old __constructor users over. Anyway, the old way is superior because
it allows multiple functions be specified. Delete the MODULE_* macros
as there's only one user of them in the entire tree.
We can also get rid of the UNKNOWN_MODULE constant because now a
module doesn't need a __module_init_ptr symbol to be classed as a
library - if a module isn't an executable, it's a library, there's no
such thing as an unknown type.
It's no longer necessary to explicitly call
__syslinux_get_ipappend_strings() from ldlinux.c because the
__constructor tag on the version in com32/lib/syslinux will take care
of invoking it before ldlinux.c32's main() is executed.
Also, since we've refactored unload_module() to now run a module's
destructors let's minimise the number of callers by deleting
load_library() and unload_library(), which were only called by the
test module in com32/elflink.
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
It would seem that running from a btrfs file system has been broken
for some time. The reason is that core/ldlinux.sys has grown
substantially, resulting in it weighing in at a whopping 88K bytes and
no longer fitting within the first 64K segment of a btrfs file system.
This huge size stems from the fact that the core links with
libcom32min.a because it requires various symbols contained within
that archive but it includes and exports *all* the symbols from
libcom32min.a. The reasoning behind this design decision is that when
a module is loaded any undefined symbols that are defined in
libcom32min.a can be resolved at runtime by the core, without the
module also needing to statically link against libcom32min.a.
Unfortunately doing this has increased the size of the core beyond
acceptable limits, such that booting from a btrfs file system no
longer works.
This commit links ldlinux.c32 against libcom32min.a so that now
ldlinux.c32 exports all the symbols contained within the
archive. Since we will always load ldlinux.c32 before any other
modules, any undefined symbols will now be resolved by ldlinux.c32
instead of the core. ldlinux.c32 isn't subject to same size
constraints as the core, e.g. fitting within 64K.
Here are the sizes in bytes before and after this commit,
Before:
33806 com32/elflink/ldlinux/ldlinux.c32
87749 core/ldlinux.sys
After:
393871 com32/elflink/ldlinux/ldlinux.c32
45516 core/ldlinux.sys
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
|
| |
|
|
|
|
|
|
|
| |
printf() is used heavily in the ldlinux code but that only displays
things on the VGA console, not on serial. Introduce eprintf(), which
gives us the benefit of format strings but will also print to both VGA
and serial.
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
|
| |
|
|
|
|
|
|
|
|
| |
Move all the code for the ADV into ldlinux so that it doesn't have any
dependencies on other modules.
We also need a way to initialize the ADV from ldlinux, so add another
vector to the comboot API.
Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
ldlinux currently requires the get_key() symbol, which means it has a
dependency on libutil_com.c32. ldlinux.c32 really should be a
standalone module so let's move get_key.c into ldlinux/.
However, moving get_key.c creates a new problem - native linux
applications such as com32/samples/keytest.c require the get_key()
symbol which is currently exported by libutil_lnx.a. To fix this, we
create a new native linux library that exports any required symbols
from ldlinux. With this change we need to update com32/Makefile so
that we build com32/elfink/ldlinux before com32/samples, and seeing as
ldlinux now has no dependencies, we may as well move it to the front
of $SUBDIRS.
Also, update elf_gen_dep.sh to skip ldlinux.c32 and not search for
dependencies for any of its unresolved symbols. This way, if anyone
inadvertently introduces a dependency that cannot be resolved by the
core, ldlinux.c32 will fail to load at runtime (which guarantees that
the newly introduced dependency won't go unnoticed).
Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
com32/elflink/modules was originally created to house ELF modules and
keep them separate from the COM32 modules as the elflink branch was
being developed. However, this has inadvertently created a maintenance
nightmare because code was copied from elsewhere in the tree into
com32/elflink/modules, resulting in duplication. Bug fixes have been
going into the original code but have not been merged onto the elflink
branch, leaving the duplicate code in com32/elflink/modules buggy.
So let's delete this directory. There really is no reason to keep ELF
and COM32 modules separate because there's no reason to need both
COM32 and ELF modules to coexist. ELF is a far superior object file
format and all modules are not emitted as ELF objects.
Now that we're outputting ELF modules we can use dynamic memory
instead of the cs_bounce bounce buffer.
This commit requires a certain amount of shuffling for some
files. quicksort.c isn't a module and belongs as part of the util
library. cli.h belongs in com32/include so that other modules can make
use of the cli code in ldlinux.c32.
All libraries are now ELF shared libraries which are only loaded to
fixup unresolved symbols for executable modules and renamed from *.a
to *.c32. This reduces the runtime memory footprint because libraries
are only loaded when needed and because every executable no longer
gets its own copy of code/data (as it would if linking with a static
library). Also, remove MINLIBOBJS from libcom32.c32 because it is
already part of libcom32min.a and we don't want to have any duplicate
symbols between the core (which links with libcom32min.a) and
libcom32.c32.
Welcome to the New World Order of ELF modules!
Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
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>
|
| |
|
|
|
|
|
| |
Catch up with recent changes to the Makefile fragments in the MCONFIG
files.
Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
|
| |
|
|
|
|
|
|
|
| |
ldlinux.c: In function ‘enter_cmdline’:
ldlinux.c:15:23: warning: unused variable ‘aux’
ldlinux.c: In function ‘ldlinux_main’:
ldlinux.c:75:5: warning: implicit declaration of function ‘kaboom’
Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
|
|
|
Currently, there is a lot of code in core/elflink that is not strictly
required for the core to function. There are benefits to keeping the
core proper as small as possible, for example so that it can fit
within a single 64K block on btrfs.
So move the following pieces of code out of core/elflink and into a
new ldlinux ELF module which will be automatically loaded by the core:
- command line interface code
- C configuration parser
- refstr implementation
- rest of the menu system
Below are some measurements of the size of ldlinux.elf before and
after this change.
Before:
text data bss dec hex filename
161328 10439 913468 1085235 108f33 core/ldlinux.elf
After:
text data bss dec hex filename
142423 10415 909500 1062338 1035c2 core/ldlinux.elf
Signed-off-by: Matt Fleming <matt.fleming@linux.intel.com>
|
