Here we add a couple of fixes and improvements for ARC processors.
All except 1 patch are already in the upstream "master" branch
and will be an essential part of GCC 11.x whenever it gets released.
The most important are first 4 patches (0005-0008) which introduce
support of full native GDB support in Linux on ARC.
And the rests are tiny, yet useful improvements.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
This fixes a defect introduced in 25162c7. The "uint" type has not
been explicitly defined here on mingw, causing compilation to fail.
Signed-off-by: Artem Panfilov <artemp@synopsys.com>
In GDB 10.x gdbserver was promoted to the top-level folder,
see https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=919adfe8409211c726c1d05b47ca59890ee648f1
Which means it is no longer a subfolder in "gdb" and so we have to
build gdbserver now exactly in the same way as normal native GDB.
One interesting detail is gdbserver doesn't need to deal with target
description in .xml so it doesn't depend on libexpat on target,
thus we need to move libexpat explicit selection from do_gdb_backend()
to its callers when building native [full] gdb as well as cross-gdb
for the host.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
[cp: support old/new layout, regenerate patches]
Signed-off-by: Chris Packham <judge.packham@gmail.com>
Changes since v0.5.0:
* Add spec files for am64x SoCs.
* Require Binutils at least version 2.37.
* Require pru-gcc to be installed.
* Remove linker scripts. Instead set memory sizes from specs.
* Activate --gc-sections linker option by default.
* The "--host=pru" configure option must be used instead of "--target=pru.
Signed-off-by: Dimitar Dimitrov <dimitar@dinux.eu>
Since 1.6.2, picolibc has had two releases - 1.7 and 1.7.1. Here are
the release notes from those versions.
1. Add __cxa_atexit implementation to 'picoexit' path as required by
C++
2. Fix lack of 'hh' support in integer-only tinystdio printf path.
3. Fix tinystdio __file flag initialization for C++ apps
1. Merge libc and libm into a single library. Having them split
doesn't offer any advantages while requiring that applications add
'-lm' to link successfully. Having them merged allows use of libm
calls from libc code.
2. Add hex float format to *printf, *scanf and strto{d,f,ld}. This is
required for C99 support.
3. Unify strto{d,f,ld} and *scanf floating point parsing code. This
ensures that the library is consistent in how floats are parsed.
4. Make strto{d,f,ld} set errno to ERANGE on overflow/underflow,
including when the result is a subnormal number.
Signed-off-by: Keith Packard <keithp@keithp.com>
Some of the installed libstdc++ header files use '#include_next' to
work around toolchain oddities that might cause loops in the
compiler. However, these also cause mistakes in locating header files
when there are multiple C libraries installed as '#include_next' often
ends up finding default C library header files.
It doesn't seem like this patch could be accepted upstream; there's a
long discussion about the use of include_next in these headers which I
cannot fully understand.
Signed-off-by: Keith Packard <keithp@keithp.com>
This commit adds the missing gcc milestones 9 and 10, so that the
helper symbols `GCC_9_or_later` and `GCC_10_or_later` can be used.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
Update linux package and add new version:
Add 5.12.4
From 5.11.6 to 5.11.21
From 5.4.105 to 5.4.119
From 4.20.9 to 4.20.17
From 4.19.180 to 4.19.190
From 4.9.261 to 4.9.268
From 4.4.261 to 4.4.268
From 3.18.134 to 3.18.139
From 3.16.82 to 3.16.85
Signed-off-by: Bensuperpc <bensuperpc@gmail.com>
strace.io currently has an expired certificate. The same files are
available via github so add that as a mirror.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
GCC11 now have -std=c++17 by default and c++17 does not allow dynamic
exception specifications.
Signed-off-by: Nik Konyuchenko <spaun2002mobile@gmail.com>
This change replicates what GLIBC 2.23 has in the string/rawmemchr.c:
// #if __GNUC_PREREQ (11, 0)
// /* Likewise GCC 11, with a different warning option. */
// DIAG_IGNORE_NEEDS_COMMENT (11, "-Wstringop-overread");
// #endif
With -Werror multiple platforms failing on the string/rawmemchr.c:40 line.
Signed-off-by: Nik Konyuchenko <spaun2002mobile@gmail.com>
Newer versions of GCC complain:
plt.c: In function 'arch_elf_add_plt_entry':
plt.c:359:3: error: '%s' directive argument is null [-Werror=format-overflow=]
359 | fprintf(stderr, "%s: failed %s(%#llx): %sn", __func__,
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
360 | name, addr, strerror(errno));
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
add a patch to avoid this error.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
As of today ARCompact (AKA "ARCv1 ISA") processors (mostly those are ARC770)
are not officially supported by upstream glibc as it adds quite some burden
on release ans support of yet another "architecture" port.
But given on ABI and ISA front ARCompact is very close to ARCv2 we may
easily retrofit its support off-the-tree, which we do here.
Also given amount of changes we need to apply there's a hope it will be easy
and straight-forward to apply the same to later versions of glibc.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Forward ported all local patches from previous glibc release 2.32, they
applied cleanly.
Signed-off-by: Hans-Christian Noren Egtvedt <hegtvedt@cisco.com>
Forward ported patches from binutils 2.35.1, refreshed to match current
sources.
Patch 0008-poison-system-directories.patch did not apply clean and had
minor adjustments to fit new binutils 2.36.1 release.
Dropped the following patches as they have been applied upstream:
- 0009-Import-patch-from-mainline-to-fix-decoding-DWARF-inf.patch
- 0010-arm-Fix-the-wrong-error-message-string-for-mve-vldr-.patch
Signed-off-by: Hans-Christian Noren Egtvedt <hegtvedt@cisco.com>
Include the gnuprumcu package in PRU cross toolchain.
Toolchain is somewhat useless without device specs and
linker scripts for the various SoCs.
Signed-off-by: Dimitar Dimitrov <dimitar@dinux.eu>
This allows building newlib-nano in addition to newlib and picolibc,
allowing users to select between C libraries within the same toolchain.
Signed-off-by: Keith Packard <keithp@keithp.com>
This adds another mode to do_gcc_core_backend that builds libstdc++
against an alternate libc implementation.
Signed-off-by: Keith Packard <keithp@keithp.com>
1. Allows picolibc's stdio to be used with c++ applications
2. Includes arm exception information in link rules
3. Adds picolibc include directory to c++ header paths.
These patches are all upstream in picolibc and included in version
1.5.
Signed-off-by: Keith Packard <keithp@keithp.com>
ftp://invisible-island.net/ncurses doesn't actually have the numbered
ncurses tarball. We've not noticed because we fall back to using the GNU
mirror which does. It's also ftp which is being deprecated.
Switch to https://invisible-mirror.net/archives/ncurses which has the
actual tarballs and provides a secure connection.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
If we are targetting an aarch64-none-elf toolchain we end up running
into a build issue in gcc/config/aarch64/driver-aarch64.c. This is
fixed in upstream gcc so just backport the patch to gcc-10.2.0
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Pull in a change from upstream GCC and one from the gcc-darwin-arm64
repo that gets an initial cross compiler building on ARM based Mac.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Update to GMP 6.2.1 for MacOS ARM support and pull in one patch
from repo that deal with a possible issue with GMP on MacOS ARM
systems.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
This commit adds support for the following Oracle products, in order
to target Oracle Linux 7.9:
Binutils 2.27-44.base.0.400
GCC 4.8.5-44.0.5
glibc 2.17-317.0.3
UEK5/u4 4.14.35-2025.400.8
Sample configuration files are provides for the following triplets:
arm-ol7u9-linux-gnueabi
arm-ol7u9-linux-gnueabihf
Signed-off-by: Egeyar Bagcioglu <egeyar.bagcioglu@oracle.com>
Signed-off-by: Jose E. Marchesi <jose.marchesi@oracle.com>
Set the origin of the Linux tarballs to www.kernel.org in order to avoid
getting an empty string in menuconfig.
Signed-off-by: Egeyar Bagcioglu <egeyar.bagcioglu@oracle.com>
[cp: use kernel.org]
Signed-off-by: Chris Packham <judge.packham@gmail.com>
When building aarch64-unknown-linux-gnu on MacOS X, aarch64-builtins.c
files doesn't build by default with clang on MacOS X. We need to pass
-std=gnu++11 when building the file for things to work with clang.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Expect that if there is a binutils 2.35.2 release these fixes will be
included in there, these are pulled out of the binutils-2_35-branch post
the 2.35.1 release.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Backport change from glibc upstream that defines RELEASE as stable
instead of release. This will at least cause the default_cflags to be
set to expected default values again.
Ref issue #1396, although the bigger issue of respecting crosstool-ng
CT_GLIBC_EXTRA_CFLAGS is most likely still not fixed.
Signed-off-by: Hans-Christian Noren Egtvedt <hegtvedt@cisco.com>
When building powerpc64+glibc-2.32 we see the following error:
[ALL ] In file included from ../sysdeps/powerpc/powerpc32/power4/multiarch/wordcopy-ppc32.c:26,
[ALL ] from ../sysdeps/powerpc/powerpc64/multiarch/wordcopy-ppc64.c:18:
[ALL ] ../string/wordcopy.c: In function '_wordcopy_fwd_aligned':
[ERROR] ../string/wordcopy.c:98:26: error: 'a1' may be used uninitialized in this function [-Werror=maybe-uninitialized]
[ALL ] 98 | ((op_t *) dstp)[0] = a1;
[ALL ] | ~~~~~~~~~~~~~~~~~~~^~~~
[ALL ] In file included from ../sysdeps/powerpc/powerpc64/power4/../../powerpc32/power4/memcopy.h:38,
[ALL ] from ../sysdeps/powerpc/powerpc64/power4/memcopy.h:1,
[ALL ] from ../string/wordcopy.c:23,
[ALL ] from ../sysdeps/powerpc/powerpc32/power4/multiarch/wordcopy-ppc32.c:26,
[ALL ] from ../sysdeps/powerpc/powerpc64/multiarch/wordcopy-ppc64.c:18:
[ALL ] ../string/wordcopy.c: In function '_wordcopy_fwd_dest_aligned':
[ERROR] ../sysdeps/generic/memcopy.h:72:61: error: 'a3' may be used uninitialized in this function [-Werror=maybe-uninitialized]
[ALL ] 72 | #define MERGE(w0, sh_1, w1, sh_2) (((w0) << (sh_1)) | ((w1) >> (sh_2)))
[ALL ] | ^~
This seems to only trigger for powerpc64. Other architectures build
cleanly. For now avoid specifying -Werror for powerpc64+glibc-2.32.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
This adds support for using picolibc instead of newlib on embedded
systems.
Signed-off-by: Keith Packard <keithp@keithp.com>
v2:
Add check for meson and ninja
Sync option default values with current picolibc defaults
Remove xtensa sys header file install as those aren't in picolibc
Forward ported patches from binutils 2.34, refreshed to match current
sources.
Patch 0008-poison-system-directories.patch did not apply clean and had
minor adjustments to fit new binutils 2.35 release.
Signed-off-by: Hans-Christian Noren Egtvedt <hegtvedt@cisco.com>
Forward ported patches from GCC 9.x, refreshed to match current sources.
Patch 0010-crystax.patch did not apply clean and had minor adjustments
to fit new GCC 10.2.0 release.
Signed-off-by: Hans-Christian Noren Egtvedt <hegtvedt@cisco.com>
Forward ported patches from GCC 9.2.0 to 9.3.0, refreshed to match
current sources.
Patch 0010-crystax.patch did not apply clean and had minor adjustments
to fit new GCC 9.3.0 release.
Signed-off-by: Hans-Christian Noren Egtvedt <hegtvedt@cisco.com>
Forward ported 0000-typedef-caddr.patch from previous release (2.31).
Dropped the 0001-Add-ARC-architecture.patch file, since ARC support is
included in glibc release 2.32.
Signed-off-by: Hans-Christian Noren Egtvedt <hegtvedt@cisco.com>
-- c6x: Add support for c6x product families to pass on to uClibC-ng
-- c6x: Fix multilib support
-- c6x: Add patch fix internal instruction error (GCC 57295)
Signed-off-by: Dan Tejada <dan.tejada@cantada.com>
The same binaries are now hosted on GitHub releases (and looking at the
homepage, that's the only download location they are offering). Use that
mirror at least as an option.
When GCC was updated to 8.4.0 the patches were moved incorrectly. Fix
this now so that they will be picked up.
Fixes: commit 31ca84ac ("Update to gcc 8.4.0")
Signed-off-by: Chris Packham <judge.packham@gmail.com>
bison, elf2flt, gcc, isl, linux, mingw-w64, strace.
Adjust checksum ordering in uclibc-ng so that it matches what the script
generates.
Signed-off-by: Alexey Neyman <stilor@att.net>
Two patches from 0.18.8.1 were dropped:
- one changing the declaration of environ is no longer needed, the
corresponding files no longer have this declaration
- one with Woe32 fixes for -O0 may need to be re-added but only after I
find what configuration breaks without it; gettext sources overwent a
massive restructuring so this patch should not be applied without
testing.
Signed-off-by: Alexey Neyman <stilor@att.net>
So again due to non-technical reasons (lack of public documentation
of ARC instruction set which we actively work on but no yet published)
we missed upstream 2.30 release.
Still the code is there, we regularly run full test-suite and are confident
in port's quality and robustness.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Currently, SourceForge is down and downloads give a 500 error. That's
not overly uncommon (even less often the case these days). Fortunately,
zlib provides another mirror on their homepage, add that as option to
the package description. (https://www.zlib.net/)
Forward ported patches from GCC 8.3.0 to 9.2.0, refreshed to match
current sources.
In patch 0012-crystax.patch, removed changing sysv4.h header file for
rs6000, since it no longer defines LINK_EH_SPEC.
Removed the following patches because they are part of upstream:
- 0018-ARC-Add-multilib-support-for-linux-targets.patch
- 0020-ARM-fix-cmse.patch
- 0021-arm-Make-arm_cmse.h-C99-compatible.patch
- 0022-ARC-Update-fma-expansions.patch
Renamed 0019-isl-0.20.patch => 0018-isl-0.20.patch.
Signed-off-by: Hans-Christian Noren Egtvedt <hegtvedt@cisco.com>
ARC port of Glibc was submitted to the mailing list recently [1]
but due to late submission it didn't make it in Glibc 2.29 release.
Hopefully by the time of next release it will be a part of upstream
release but for now we have to use off-the-tree patch.
Still it's proven to work internally as well as its test-suite
shows brilliant results as might be seen from [1]:
------------------>8-------------------
Summary of test results:
24 FAIL
5124 PASS
27 UNSUPPORTED
19 XFAIL
------------------>8-------------------
Moreover ARC's Glibc port is known to work in Buildroot, OpenEmbedded
and even Automotive Grade Linux distro so we should be good having
this patch for Glibc.
BTW the patch itself is a copy of the one I use in OE, see [2].
[1] https://sourceware.org/ml/libc-alpha/2018-12/msg00678.html
[2] https://github.com/foss-for-synopsys-dwc-arc-processors/meta-synopsys/blob/master/recipes-core/glibc/files/0031-Add-ARC-architecture.patch
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Older GCCs (e.g. GCC 4.4 in CentOS 6) chokes on having two identical
typedefs like this:
typedef void *p;
typedef void *p;
The event-loop.h header does not have a guard against double-inclusion.
Not the best solution, but it works and eventually 7.12 will fall into
obsolescence anyway.
Signed-off-by: Alexey Neyman <stilor@att.net>
- Force building make as a companion tool if host make is older than
4.0 (CentOS 7 currently has 3.82)
- Disable 2.29 as a choice if host python is older than 3.4
(CentOS 7 has 2.6 unless python from EPEL is installed)
- Python2 emits its version information to STDERR. Ugh.
While there, also use the detected host Python for GDB configuration.
Signed-off-by: Alexey Neyman <stilor@att.net>
Had to update the following patches so they'd apply:
0006-libgfortran-missing-include.patch
0010-crystax.patch
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
When merging new package framework, I missed that master had 7.8.2
imported in place of 7.8.1. Noone complained about this downgrade...
Signed-off-by: Alexey Neyman <stilor@att.net>
... in uClibc and glibc.
Fixes#681.
While here, relocate additional "sources" for uClibc/binutils into packages/
directory.
Signed-off-by: Alexey Neyman <stilor@att.net>
This required some rework of the libc selection, as moxiebox is a layer on
top of another libc - newlib.
Also, moxiebox'es host VM (`sandbox`) needs a libcrypto on the host. We will
not have it if we're cross-compiling a canadian cross. Fortunately, all moxiebox
needs from libcrypto is SHA256, and it already includes a standalone implementation
of SHA256 in its runtime. Provide a little wrapper that allows moxiebox use
that implementation for the host binary, too.
Also, automate collecting/printing the list of all packages in a given category
(e.g. LIBC or COMP_TOOLS), generate a list of all Kconfig symbols for a given
category.
Signed-off-by: Alexey Neyman <stilor@att.net>
There are some fixes to the cmse code in mainline gcc that we need to
build an embedded toolchain targetting Cortex-M cpus that support the
v8m extensions.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
gcc-8.1 for xtensa miscompiles uClibc dynamic linker due to gcc PR
target/65416. The build completes successfully, but the binary is
non-functional because the following fragment in the _dl_get_ready_to_run
in ld-uClibc.so overwrites register spill area on stack causing register
corruption in the previous call frame and a subsequent crash:
419f: f0c1b2 addi a11, a1, -16
41a2: 1ba9 s32i.n a10, a11, 4
41a4: 0bc9 s32i.n a12, a11, 0
41a6: 5127f2 l32i a15, a7, 0x144
41a9: 1765b2 s32i a11, a5, 92
41ac: 4e2782 l32i a8, a7, 0x138
41af: 146af2 s32i a15, a10, 80
41b2: 001b10 movsp a1, a11
The crash terminates the init process and causes kernel panic.
The fix prevents reordering of movsp opcode and any access to the stack
frame memory and is applicable to all existing gcc versions.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
We never bothered to build multilibbed Linux toolchains for ARC
and so we ended-up with no support of multilib for arc*-*-linux-*
targets in GCC. That is now fixed in upstream by the following commit:
https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=0eacfbcb2bf1834294f468a2bb41fe5d5e8d3883
Once the fix mentioned above is applied on top of GCC 8.1.0 we may easily build
multilibbed uClibc toolchain for ARC.
Note sice Glibc port for ARC is still in review process we cannot build any Glibc
toolchains for ARC with pure upstream components, for that we eiter need Glibc
sources from our Github repo or apply a set of patches on top of Glibc 2.27 - we'll
try to address this later though.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
hs4x and hs4xd were mistakenly mentioned in MULTILIB_OPTION/DIRNAMES
before real support of those CPUs was accepted in upstream.
That breaks multilib toolchains building becase we're trying to compile
for not yet known CPU.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Without this fix we're getting the following error on attempt to build
Linux kernel:
--------------------------------->8---------------------------------
xfrm4_mode_tunnel.s: Assembler messages:
xfrm4_mode_tunnel.s:188: Error: operand out of range (128 is not between -
128 and 127)
--------------------------------->8---------------------------------
The fix is taken from
d8d716f49c
and it is supposed to be merged in upstream GCC sources soonish.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
- Incompatible function type for ifunc alias
- Multiple statements macro expansion in strftime
- if_nametoindex size checking
Signed-off-by: Alexey Neyman <stilor@att.net>
