Florian Weimer [Tue, 30 Aug 2022 11:30:03 +0000 (13:30 +0200)]
resolv: Fix building tst-resolv-invalid-cname for earlier C standards
This fixes this compiler error:
tst-resolv-invalid-cname.c: In function ‘test_mode_to_string’:
tst-resolv-invalid-cname.c:164:10: error: label at end of compound statement
case test_mode_num:
^~~~~~~~~~~~~
The a583b6add407c17cd change did not handle large messages that
would require a heap allocation correctly, where the message itself
is not take in consideration.
This patch fixes it and extend the tst-syslog to check for large
messages as well.
Stefan Liebler [Thu, 25 Aug 2022 10:17:48 +0000 (12:17 +0200)]
S390: Always use svc 0
On s390x syscalls are triggered by svc instruction. One can
pass the syscall number encoded in the instruction "svc 123"
or by storing it in r1:
lghi r1,123
svc 0
If the syscall number is encoded in the instruction, this can
cause broken syscall restarts. Therefore this patch is now just
passing the syscall number in r1.
See also kernel-commit:
"s390/signal: switch to using vdso for sigreturn and syscall restart"
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/arch/s390/[%e2%80%a6]call.c?h=v6.0-rc1&id=df29a7440c4b5c65765c8f60396b3b13063e24e9
Florian Weimer [Tue, 30 Aug 2022 08:02:49 +0000 (10:02 +0200)]
nss_dns: Rewrite _nss_dns_gethostbyname4_r using current interfaces
Introduce struct alloc_buffer to this function, and use it and
struct ns_rr_cursor in gaih_getanswer_slice. Adjust gaih_getanswer
and gaih_getanswer_noaaaa accordingly.
Florian Weimer [Tue, 30 Aug 2022 08:02:49 +0000 (10:02 +0200)]
nss_dns: Rewrite _nss_dns_gethostbyaddr2_r and getanswer_ptr
The simplification takes advantage of the split from getanswer_r.
It fixes various aliases issues, and optimizes NSS buffer usage.
The new DNS packet parsing helpers are used, too.
Florian Weimer [Tue, 30 Aug 2022 08:02:49 +0000 (10:02 +0200)]
resolv: Add DNS packet parsing helpers geared towards wire format
The public parser functions around the ns_rr record type produce
textual domain names, but usually, this is not what we need while
parsing DNS packets within glibc. This commit adds two new helper
functions, __ns_rr_cursor_init and __ns_rr_cursor_next, for writing
packet parsers, and struct ns_rr_cursor, struct ns_rr_wire as
supporting types.
In theory, it is possible to avoid copying the owner name
into the rname field in __ns_rr_cursor_next, but this would need
more functions that work on compressed names.
Eventually, __res_context_send could be enhanced to preserve the
result of the packet parsing that is necessary for matching the
incoming UDP packets, so that this works does not have to be done
twice.
Florian Weimer [Tue, 30 Aug 2022 08:02:49 +0000 (10:02 +0200)]
resolv: Add the __ns_samebinaryname function
During packet parsing, only the binary name is available. If the name
equality check is performed before conversion to text, we can sometimes
skip the last step.
Florian Weimer [Tue, 30 Aug 2022 08:02:49 +0000 (10:02 +0200)]
resolv: Add internal __res_binary_hnok function
During package parsing, only the binary representation is available,
and it is convenient to check that directly for conformance with host
name requirements.
nscd: Fix netlink cache invalidation if epoll is used [BZ #29415]
Processes cache network interface information such as whether IPv4 or IPv6
are enabled. This is only checked again if the "netlink timestamp" provided
by nscd changed, which is triggered by netlink socket activity.
However, in the epoll handler for the netlink socket, it was missed to
assign the new timestamp to the nscd database. The handler for plain poll
did that properly, copy that over.
This bug caused that e.g. processes which started before network
configuration got unusuable addresses from getaddrinfo, like IPv6 only even
though only IPv4 is available:
https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/issues/1041
It's a bit hard to reproduce, so I verified this by checking the timestamp
on calls to __check_pf manually. Without this patch it's stuck at 1, now
it's increasing on network changes as expected.
Andreas Schwab [Mon, 29 Aug 2022 13:05:40 +0000 (15:05 +0200)]
Add test for bug 29530
This tests for a bug that was introduced in commit edc1686af0 ("vfprintf:
Reuse work_buffer in group_number") and fixed as a side effect of commit 6caddd34bd ("Remove most vfprintf width/precision-dependent allocations
(bug 14231, bug 26211).").
Makeconfig: Set pie-ccflag to -fPIE by default [BZ# 29514]
We should default to the larger code model, in order to support
larger applications built with -static -pie. This should be
consistent with pic-ccflag, which defaults to -fPIC.
Remove the now redundant override from sysdeps/sparc/Makefile.
Note that -fno-pie and -fno-PIE have the same effect.
Reviewed-by: Carlos O'Donell <carlos@redhat.com> Tested-by: Carlos O'Donell <carlos@redhat.com>
Samuel Thibault [Sun, 28 Aug 2022 23:42:47 +0000 (01:42 +0200)]
hurd: Fix vm_size_t incoherencies
In gnumach, 3e1702a65fb3 ("add rpc_versions for vm types") changed the type
of vm_size_t, making it always a unsigned long. This made it incompatible on
x86 with size_t. Even if we may want to revert it to unsigned int, it's
better to fix the types of parameters according to the .defs files.
Samuel Thibault [Sat, 27 Aug 2022 11:52:46 +0000 (13:52 +0200)]
mach: Fix incoherency between perror and strerror
08d2024b4167 ("string: Simplify strerror_r") inadvertently made
__strerror_r print unknown error system in decimal while the original
code was printing it in hexadecimal. perror was kept printing in
hexadecimal in 725eeb4af14c ("string: Use tls-internal on strerror_l"),
let us keep both coherent.
This also fixes a duplicate ':'
Spotted by the libunistring testsuite test-perror2
Florian Weimer [Fri, 26 Aug 2022 19:15:43 +0000 (21:15 +0200)]
elf: Call __libc_early_init for reused namespaces (bug 29528)
libc_map is never reset to NULL, neither during dlclose nor on a
dlopen call which reuses the namespace structure. As a result, if a
namespace is reused, its libc is not initialized properly. The most
visible result is a crash in the <ctype.h> functions.
To prevent similar bugs on namespace reuse from surfacing,
unconditionally initialize the chosen namespace to zero using memset.
Szabolcs Nagy [Mon, 22 Aug 2022 13:21:53 +0000 (14:21 +0100)]
csu: Change start code license to have link exception
The start code can get linked into dynamic linked executables where
LGPL would require shipping the source or linkable binaries when the
executable is distributed.
On some targets the license exception was missing in start.S (which
is compiled into crt1.o and Scrt1.o which may end up linked into PDE
and PIE binaries).
I did not review what other code may end up in executables, just
fixed the start.S license inconsistency across targets.
Florian Weimer [Thu, 25 Aug 2022 14:34:20 +0000 (16:34 +0200)]
s390: Move hwcaps/platform names out of _rtld_global_ro
Changes to these arrays are often backported to stable releases,
but additions to these arrays shift the offsets of the following
_rltd_global_ro members, thus breaking the GLIBC_PRIVATE ABI.
Obviously, this change is itself an internal ABI break, but at least
it will avoid further ABI breaks going forward.
Once this change hits the release branches, it will require relinking
of all statically linked applications before static dlopen works
again, for the majority of updates on release branches: The NEWS file
is regularly updated with bug references, so the __libc_early_init
suffix changes, and static dlopen cannot find the function anymore.
While this ABI check is still technically correct (we do require
rebuilding & relinking after glibc updates to keep static dlopen
working), it is too drastic for stable release branches.
Florian Weimer [Wed, 24 Aug 2022 15:35:36 +0000 (17:35 +0200)]
Detect ld.so and libc.so version inconsistency during startup
The files NEWS, include/link.h, and sysdeps/generic/ldsodefs.h
contribute to the version fingerprint used for detection. The
fingerprint can be further refined using the --with-extra-version-id
configure argument.
_dl_call_libc_early_init is replaced with _dl_lookup_libc_early_init.
The new function is used store a pointer to libc.so's
__libc_early_init function in the libc_map_early_init member of the
ld.so namespace structure. This function pointer can then be called
directly, so the separate invocation function is no longer needed.
The versioned symbol lookup needs the symbol versioning data
structures, so the initialization of libc_map and libc_map_early_init
is now done from _dl_check_map_versions, after this information
becomes available. (_dl_map_object_from_fd does not set this up
in time, so the initialization code had to be moved from there.)
This means that the separate initialization code can be removed from
dl_main because _dl_check_map_versions covers all maps, including
the initial executable loaded by the kernel. The lookup still happens
before relocation and the invocation of IFUNC resolvers, so IFUNC
resolvers are protected from ABI mismatch.
The __libc_early_init function pointer is not protected because
so little code runs between the pointer write and the invocation
(only dynamic linker code and IFUNC resolvers).
Paul Eggert [Wed, 24 Aug 2022 04:16:39 +0000 (21:16 -0700)]
Merge _GL_UNUSED C23 patch from Gnulib
* posix/getopt.c (_getopt_initialize):
* sysdeps/posix/tempname.c (try_dir, try_nocreate):
Put _GL_UNUSED before args instead of after.
This makes no difference for glibc.
It is needed for Gnulib when being compiled on
non-GCC C23 compilers.
Samuel Thibault [Mon, 22 Aug 2022 20:34:31 +0000 (22:34 +0200)]
hurd: Fix starting static binaries with stack protection enabled
gcc introduces gs:0x14 accesses in most functions, so we need some tcbhead
to be ready very early during initialization. This configures a static area
which can be referenced by various protected functions, until proper TLS is
set up.
Joseph Myers [Mon, 22 Aug 2022 14:16:57 +0000 (14:16 +0000)]
Add AGROUP from Linux 5.19 to sys/acct.h, remove Alpha version (bug 29502)
Linux 5.19 adds a new accounting flag AGROUP; add it to the
enumeration in sys/acct.h.
This shows up that the Alpha-specific variant of this header has a
different set of constants and struct acct, which appear to be the
constants and structure layout from Linux 2.0. These were changed
some time between Linux 2.0 and Linux 2.2; I see no evidence of an
Alpha-specific layout or set of constants, but haven't checked the
detailed Linux kernel history between those versions. Rather, it
looks like tha Alpha-specific header was originally needed because of
the use of types in the kernel structure (such as uid_t and gid_t)
that had different sizes on Alpha, and when glibc was updated for
changes to the structure and constants in the kernel
1998-10-02 Andreas Jaeger <aj@arthur.rhein-neckar.de>
* sysdeps/unix/sysv/linux/sys/acct.h: Bring in sync with current
linux 2.1 version.
that simply omitted to do anything about the Alpha version.
Thus, remove the Alpha version in order to get the updated definitions
into use on Alpha, as I don't think the interfaces are actually
different for Alpha with any kernel version supported by glibc.
Tested for x86_64, and with build-many-glibcs.py for alpha-linux-gnu.
Joseph Myers [Thu, 18 Aug 2022 13:16:56 +0000 (13:16 +0000)]
Use binutils 2.39 branch in build-many-glibcs.py
This patch makes build-many-glibcs.py use binutils 2.39 branch.
Tested with build-many-glibcs.py (compilers and glibcs builds). Note:
binutils 2.39 shows the same failures for i686-linux-gnu-no-pie,
x86_64-linux-gnu-no-pie and x86_64-linux-gnu-x32-no-pie building the
glibc testsuite as binutils mainline does.
Stefan Liebler [Thu, 18 Aug 2022 07:10:48 +0000 (09:10 +0200)]
S390: Fix werror=unused-variable in ifunc-impl-list.c.
If the architecture level set is high enough, no IFUNCs are used at
all and the variable i would be unused. Then the build fails with:
../sysdeps/s390/multiarch/ifunc-impl-list.c: In function ‘__libc_ifunc_impl_list’:
../sysdeps/s390/multiarch/ifunc-impl-list.c:76:10: error: unused variable ‘i’ [-Werror=unused-variable]
76 | size_t i = max;
| ^
cc1: all warnings being treated as errors
Ensure calculations happen with desired rounding mode in y1lf128
math/test-float128-y1 fails on x86_64 and ppc64el with gcc 12 and -O3,
because code inside a block guarded by SET_RESTORE_ROUNDL is being moved
after the rounding mode has been restored. Use math_force_eval to
prevent this (and insert some math_opt_barrier calls to prevent code
from being moved before the rounding mode is set).
Florian Weimer [Tue, 16 Aug 2022 07:25:23 +0000 (09:25 +0200)]
Linux: Fix enum fsconfig_command detection in <sys/mount.h>
The #ifdef FSOPEN_CLOEXEC check did not work because the macro
was always defined in this header prior to the check, so that
the <linux/mount.h> contents did not matter.
Florian Weimer [Tue, 16 Aug 2022 06:27:50 +0000 (08:27 +0200)]
elf: Run tst-audit-tlsdesc, tst-audit-tlsdesc-dlopen everywhere
The test is valid for all TLS models, but we want to make a reasonable
effort to test the GNU2 model specifically. For example, aarch64
defaults to GNU2, but does not have -mtls-dialect=gnu2, and the test
was not run there.
Florian Weimer [Mon, 15 Aug 2022 14:45:40 +0000 (16:45 +0200)]
malloc: Do not use MAP_NORESERVE to allocate heap segments
Address space for heap segments is reserved in a mmap call with
MAP_ANONYMOUS | MAP_PRIVATE and protection flags PROT_NONE. This
reservation does not count against the RSS limit of the process or
system. Backing memory is allocated using mprotect in alloc_new_heap
and grow_heap, and at this point, the allocator expects the kernel
to provide memory (subject to memory overcommit).
The SIGSEGV that might generate due to MAP_NORESERVE (according to
the mmap manual page) does not seem to occur in practice, it's always
SIGKILL from the OOM killer. Even if there is a way that SIGSEGV
could be generated, it is confusing to applications that this only
happens for secondary heaps, not for large mmap-based allocations,
and not for the main arena.
Now that kernel exports linux/mount.h and includes it on linux/fs.h,
its definitions might clash with glibc exports sys/mount.h. To avoid
the need to rearrange the Linux header to be always after glibc one,
the glibc sys/mount.h is changed to:
1. Undefine the macros also used as enum constants. This covers prior
inclusion of <linux/mount.h> (for instance MS_RDONLY).
2. Include <linux/mount.h> based on the usual __has_include check
(needs to use __has_include ("linux/mount.h") to paper over GCC
bugs.
3. Define enum fsconfig_command only if FSOPEN_CLOEXEC is not defined.
(FSOPEN_CLOEXEC should be a very close proxy.)
4. Define struct mount_attr if MOUNT_ATTR_SIZE_VER0 is not defined.
(Added in the same commit on the Linux side.)
This patch also adds some tests to check if including linux/fs.h and
linux/mount.h after and before sys/mount.h does work.
Wilco Dijkstra [Thu, 11 Aug 2022 15:47:45 +0000 (16:47 +0100)]
libio: Improve performance of IO locks
Improve performance of recursive IO locks by adding a fast path for
the single-threaded case. To reduce the number of memory accesses for
locking/unlocking, only increment the recursion counter if the lock
is already taken.
On Neoverse V1, a microbenchmark with many small freads improved by
2.9x. Multithreaded performance improved by 2%.
So far this test checks if pidfd_open-syscall is supported,
which was introduced with linux 5.3.
The process_madvise-syscall was introduced with linux 5.10.
Thus you'll get FAILs if you are running a kernel in between.
This patch adds a check if the first process_madvise-syscall
returns ENOSYS and in this case will fail with UNSUPPORTED. Reviewed-by: Florian Weimer <fweimer@redhat.com>
Noah Goldstein [Mon, 8 Aug 2022 03:26:21 +0000 (11:26 +0800)]
x86: Fix `#define STRCPY` guard in strcpy-sse2.S
`#ifndef STPCPY` is incorrect for checking if `STRCPY` is already
defined. It doesn't end up mattering as the whole check is
guarded by `#if IS_IN (libc)` but is incorrect none the less.
Noah Goldstein [Mon, 8 Aug 2022 03:26:22 +0000 (11:26 +0800)]
elf: Replace `strcpy` call with `memcpy` [BZ #29454]
GCC normally does this optimization for us in
strlen_pass::handle_builtin_strcpy but only for optimized
build. To avoid needing to include strcpy.S in the rtld build to
support the debug build, just do the optimization by hand.
The older libc versions are obsolete for over twenty years now.
This patch removes the special flags for libc5 and libc4 and assumes
that all libraries cached are libc6 compatible and use FLAG_ELF_LIBC6.
Checked with a build for all affected architectures.
Florian Weimer [Thu, 4 Aug 2022 04:17:50 +0000 (06:17 +0200)]
Linux: Remove exit system call from _exit
exit only terminates the current thread, not the whole process, so it
is the wrong fallback system call in this context. All supported
Linux versions implement the exit_group system call anyway.
Joseph Myers [Wed, 3 Aug 2022 16:31:58 +0000 (16:31 +0000)]
Update kernel version to 5.19 in header constant tests
This patch updates the kernel version in the tests tst-mman-consts.py,
tst-mount-consts.py and tst-pidfd-consts.py to 5.18. (There are no
new constants covered by these tests in 5.19, or in 5.17 or 5.18 in
the case of tst-mount-consts.py that previously used version 5.16,
that need any other header changes.)
Florian Weimer [Wed, 3 Aug 2022 09:41:53 +0000 (11:41 +0200)]
assert: Do not use stderr in libc-internal assert
Redirect internal assertion failures to __libc_assert_fail, based on
based on __libc_message, which writes directly to STDERR_FILENO
and calls abort. Also disable message translation and reword the
error message slightly (adjusting stdlib/tst-bz20544 accordingly).
As a result of these changes, malloc no longer needs its own
redefinition of __assert_fail.
__libc_assert_fail needs to be stubbed out during rtld dependency
analysis because the rtld rebuilds turn __libc_assert_fail into
__assert_fail, which is unconditionally provided by elf/dl-minimal.c.
This change is not possible for the public assert macro and its
__assert_fail function because POSIX requires that the diagnostic
is written to stderr.
Florian Weimer [Wed, 3 Aug 2022 09:41:53 +0000 (11:41 +0200)]
nptl: Remove uses of assert_perror
__pthread_sigmask cannot actually fail with valid pointer arguments
(it would need a really broken seccomp filter), and we do not check
for errors elsewhere.
Florian Weimer [Wed, 3 Aug 2022 09:41:53 +0000 (11:41 +0200)]
stdio: Clean up __libc_message after unconditional abort
Since commit ec2c1fcefb200c6cb7e09553f3c6af8815013d83 ("malloc:
Abort on heap corruption, without a backtrace [BZ #21754]"),
__libc_message always terminates the process. Since commit a289ea09ea843ced6e5277c2f2e63c357bc7f9a3 ("Do not print backtraces
on fatal glibc errors"), the backtrace facility has been removed.
Therefore, remove enum __libc_message_action and the action
argument of __libc_message, and mark __libc_message as _No_return.
Joseph Myers [Tue, 2 Aug 2022 21:05:07 +0000 (21:05 +0000)]
Update syscall lists for Linux 5.19
Linux 5.19 has no new syscalls, but enables memfd_secret in the uapi
headers for RISC-V. Update the version number in syscall-names.list
to reflect that it is still current for 5.19 and regenerate the
arch-syscall.h headers with build-many-glibcs.py update-syscalls.
Arjun Shankar [Tue, 2 Aug 2022 09:10:25 +0000 (11:10 +0200)]
socket: Check lengths before advancing pointer in CMSG_NXTHDR
The inline and library functions that the CMSG_NXTHDR macro may expand
to increment the pointer to the header before checking the stride of
the increment against available space. Since C only allows incrementing
pointers to one past the end of an array, the increment must be done
after a length check. This commit fixes that and includes a regression
test for CMSG_FIRSTHDR and CMSG_NXTHDR.
The Linux, Hurd, and generic headers are all changed.
Tested on Linux on armv7hl, i686, x86_64, aarch64, ppc64le, and s390x.
Samuel Thibault [Thu, 28 Jul 2022 20:01:49 +0000 (22:01 +0200)]
htl: Let pthread_self and cancellability called early
When applications redirect some functions they might get called before
libpthread is fully initialized. They may still expected pthread_self
and cancellable functions to work, so cope with such calls in that
situation.
It uses the bitmask with rejection [1], which calculates a mask
being the lowest power of two bounding the request upper bound,
successively queries new random values, and rejects values
outside the requested range.
Performance-wise, there is no much gain in trying to conserve
bits since arc4random is wrapper on getrandom syscall. It should
be cheaper to just query a uint32_t value. The algorithm also
avoids modulo and divide operations, which might be costly
depending of the architecture.
Florian Weimer [Mon, 1 Aug 2022 13:49:07 +0000 (15:49 +0200)]
malloc: Use __getrandom_nocancel during tcache initiailization
Cancellation currently cannot happen at this point because dlopen
as used by the unwind link always performs additional allocations
for libgcc_s.so.1, even if it has been loaded already as a dependency
of the main executable. But it seems prudent not to rely on this
quirk.
Tom Honermann [Sun, 24 Jul 2022 05:11:43 +0000 (01:11 -0400)]
stdlib: Suppress gcc diagnostic that char8_t is a keyword in C++20 in uchar.h.
gcc 13 issues the following diagnostic for the uchar.h header when the
-Wc++20-compat option is enabled in C++ modes that do not enable char8_t
as a builtin type (C++17 and earlier by default; subject to _GNU_SOURCE
and the gcc -f[no-]char8_t option).
warning: identifier ‘char8_t’ is a keyword in C++20 [-Wc++20-compat]
This change modifies the uchar.h header to suppress the diagnostic through
the use of '#pragma GCC diagnostic' directives for gcc 10 and later (the
-Wc++20-compat option was added in gcc version 10). Unfortunately, a bug
in gcc currently prevents those directives from having the intended effect
as reported at https://gcc.gnu.org/PR106423. A patch for that issue has
been submitted and is available in the email thread archive linked below.
https://gcc.gnu.org/pipermail/gcc-patches/2022-July/598736.html