* Move the logic to build_cross_tools, as some scripts use this
directly.
* Use gnu89 standard instead of disabling specific warnings.
* Reduce optimization level used to compile GCC, to try and stave
off possible bugs.
* Explain why 32-bit is used for GCC 2. (With some hacks,
GCC 2 compiled in 64-bit mode can be used to build an image
that does get to the desktop, though.)
Tested with GCC 14.2.
* Use LITTLE_ENDIAN instead of __LITTLE_ENDIAN (LITTLE_ENDIAN is what
POSIX now specifies and everyone seems to define - NetBSD also
defines _LITTLE_ENDIAN, but not __LITTLE_ENDIAN)
* file descriptors greater than 2 must be explicitly passed to a child
on NetBSD. POSIX leaves that unspecified, see
https://pubs.opengroup.org/onlinepubs/9799919799/utilities/V3_chap02.html#exec
* system library functions get declared with visibility "default" on
NetBSD, so a "hidden" visibility attribute is ignored on a
redeclaration. Work around by putting the functions in a namespace -
as they are extern "C", they still end up in the object file without
the namespace qualifier.
* remove the cast on the NULL pointer on the funopen call - it's not
needed and the types are different between FreeBSD (fpos_t) and
NetBSD/OpenBSD (off_t); also trying to change that upstream
https://github.com/openSUSE/libsolv/pull/603
Change-Id: I6ab7f3c74d18960d7589b403a1c219cdac85a453
Reviewed-on: https://review.haiku-os.org/c/haiku/+/10133
Tested-by: Commit checker robot <[email protected]>
Reviewed-by: Adrien Destugues <[email protected]>
* "egrep: warning: egrep is obsolescent; using grep -E"
* egrep on haiku is a shell script calling grep -E
* given the warnings in the latest egrep, lets get
ahead of the curve and migrate to grep -E
Change-Id: I0fc2b936c7a9c52f87caf97d8e9ac42516fee014
Reviewed-on: https://review.haiku-os.org/c/haiku/+/5718
Reviewed-by: Adrien Destugues <[email protected]>
Tested-by: Commit checker robot <[email protected]>
Recently we introduced Spleen as a larger font for KDL for high
resolution displays. However, it looks quite different from our original
font. I designed an enlarged version of our original font (designed by
Brian J. Swetland and also found in NewOS, LK and Fuchsia) to use
instead.
Change-Id: I10872e407d45b906f8b6c7ba44b2f993dcd19bff
Reviewed-on: https://review.haiku-os.org/c/haiku/+/5382
Tested-by: Commit checker robot <[email protected]>
Reviewed-by: Adrien Destugues <[email protected]>
CLang fails to build ISL otherwise, and if you are setting up
cross compiling, you should have a modern enough compiler
Fixes#16434
Change-Id: Ieb4a33015899896ea3bb2d1abc59b10fe0b4b3f2
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3730
Reviewed-by: Fredrik Holmqvist <[email protected]>
* Enable TLS for riscv64. For now select a dummy
implementation.
* RISC-V has a register (tp) dedicated to TLS
* All the "desktop-like" RV64 chipsets implement GC
* Same ISA we currently build for... but calls it out
in-case defaults change.
Change-Id: I623b6e5c309b6a6e80ec378e456b2335c561269d
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3632
Reviewed-by: Adrien Destugues <[email protected]>
* This was originally added circa 2012. via hrev43954
* Our toolchain appears to work fine now with explicit
multilib enabled.
Change-Id: I1f5d55c63a6b3b6ab1c4db95da3d14d5ef2da0df
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3315
Reviewed-by: Jérôme Duval <[email protected]>
Previously, all packages on the image, both activated and unactivated,
were in $systemPackages, so if a package actually in /system/packages
had a dependency in the Installer optional package area, the dependency
checking would see no problem but the generated image would be broken.
Now, packages going into /system/packages are treated separately
from those going elsewhere, and error checks added so the above
condition will not occur.
(I ran into this while working on the build-package upgrade
in the following commits.)
We always want to use the shared one from the syslibs package,
never the shared one from this build.
Part of #14842: after this change, the ICU build now fails with
"cannot find libstdc++" instead of an invalid symbol.
This just stops the errors from occuring rather than trying to
rebuild the files at all. This is much cleaner, and solves
a few cases that the other method did not.
Our implementation of it now behaves properly after the last commit.
Change-Id: I6bebc91ae0f9512ea07ad6a7a4ccea9ee758e01b
Reviewed-on: https://review.haiku-os.org/c/908
Reviewed-by: waddlesplash <[email protected]>
It seems that not all Linux distributions ship an EFI-enabled
cdrtools (i.e. mkisofs takes -e option), Arch being one that
does not.
So instead, we now use xorriso universally, which is
as (or more, in most cases) widely available, and supports
emulating mkisofs with the EFI commands universally.
This also has the added benefit that we can drop genisoimage
support altogether.
When HEAD is tagged, the output will be identical to what it was
before (the latest hrev tag and nothing else.) When HEAD is not tagged,
and the most recent tag is further back, we now use a format like this:
hrevXXXXX+N(+dirty)
... where N is the number of commits since hrevXXXXX, and +dirty is added
if the working tree is dirty. This is significantly shorter than the
previous model (as it does not have the Git revision.)
Fixes#14445.
Change-Id: Ide7f66cf0ac1c1f05402afc52b6be3b68b66d6dc
Reviewed-on: https://review.haiku-os.org/566
Reviewed-by: waddlesplash <[email protected]>
It is just straight up broken on most systems without this, as
the same "cast specifies signature type" appears there too.
Exactly why this fixes the problem does not appear to be any more known
than it was when this was first instated in 2012.
Previously they just silently continued, which meant that if part of the
cross-tools build failed, you would have to scroll back pretty far to
see what the failure was.
Previously --update was kind of a hack, as it just dumped all
the variables from BuildConfig back into ./configure and attempted
to rerun from there.
Instead, now we store all of the variables and environs configure
was invoked with in the top of BuildConfig and re-call it using them.
This is much more robust, as configure will actually recompute
all of the values from the BuildConfig.
As a side effect, if configure was originally invoked to build crosstools,
it would have done so again, so instead add a check to see if they
exist, and skip them if they do. It's easy enough to delete the
directory if you want to force a crosstools rebuild.
Briefly discussed with mmlr on IRC.
* Previously, the cross-compiler would generate code that doesn't
run on Haiku, particularly where TLS is concerned. It also ended
up with a c++config.h header incompatible with the version in
the native compiler.
* Now possible to correctly cross-compile rust for Haiku.
Revert "repo rework: Remove stubs; Breaks repo compat."
Revert "repo rework: Remove need for repos to be self-aware"
This reverts commit a2b2f4d642.
This reverts commit 602076ef82.
This reverts commit 5ffaf72c8a.
These changes break the build on Haiku and the ability to create repo
mirrors, for the lack of a replacement for the URL (an UUID was evoked
on the mailing lists, but not implemented).
We are due for a release soon, please don't break the build.
* See #12917 for details.
* Squashed to one commit to make revert easy if we
run into any issues.
* pkgman is now pre-attached to the 'current' repo
version within nightly images so they can be updated
by default.
* This shouldn't impact us keeping older sets of package
versions by commit hash for building older hrevs.
* There are XXX stubs with "Kill me". These will need
to be dropped after users are given sufficent time to
upgrade. We're dropping a previously required field (url)
so making this a slowish roll out.
* Makes the repos a lot less restrictive which should
help PM package building automation be a bit easier.
* Once this stuff smooths out, we'll add UUID's to the
repo definitions for duplicate repo detection.
from a stock GDB source tree.
I was getting tired of doing it manually.
Note the prefix is not exactly the same as the buildtools since the
machine triplet is different, but it's not an issue.
Tested with arm and ppc with GDB 7.8. Might need patching gdbTarget for other archs,
as stock GDB doesn't know about Haiku.
SED4906