the workaround/fallback code for supporting O_PATH file descriptors
when the kernel lacks support for performing these operations on them
caused EBADF to get replaced by ENOENT (due to missing entry in
/proc/self/fd). this is unlikely to affect real-world code (calls that
might yield EBADF are generally unsafe, especially in library code)
but it was breaking some test cases.
the fix I've applied is something of a tradeoff: it adds one syscall
to these operations on kernels where the workaround is needed. the
alternative would be to catch ENOENT from the /proc lookup and
translate it to EBADF, but I want to avoid doing that in the interest
of not touching/depending on /proc at all in these functions as long
as the kernel correctly supports the operations. this is following the
general principle of isolating hacks to code paths that are taken on
broken systems, and keeping the code for correct systems completely
hack-free.
the ABI allows the callee to clobber stack slots that correspond to
arguments passed in registers, so the caller must adjust the stack
pointer to reserve space appropriately. prior to this fix, the argv
array was possibly clobbered by dynamic linker code before passing
control to the main program.
our getcwd already (as an extension) supports allocation of a buffer
when the buffer argument is a null pointer, so there's no need to
duplicate the allocation logic in this wrapper function. duplicating
it is actually harmful in that it doubles the stack usage from
PATH_MAX to 2*PATH_MAX.
the wildcard function in GNU make includes dangling symlinks; if any
exist under the .git directory, they would get added as dependencies,
causing make to exit with an error due to lacking a rule to build the
missing file.
as far as I can tell, git operations which should force version.h to
be rebuilt must all touch the mtime of the top-level .git directory.
previously this flag was defined and accepted as a no-op, possibly
breaking some software that uses it. given the choice to remove the
definition and possibly break applications that were already working,
or simply implement the feature, the latter turned out to be easy
enough to make the decision easy.
in the case where the FNM_PATHNAME flag is also set, this
implementation is clean and essentially optimal. otherwise, it's an
inefficient "brute force" implementation. at some point, when cleaning
up and refactoring this code, I may add a more direct code path for
handling FNM_LEADING_DIR in the non-FNM_PATHNAME case, but at this
point my main interest is avoiding introducing new bugs in the code
that implements the standard fnmatch features specified by POSIX.
this is still experimental and subject to change. for git checkouts,
an attempt is made to record the exact revision to aid in bug reports
and debugging. no version information is recorded in the static libc.a
or binaries it's linked into.
the FNM_PATHNAME logic for advancing by /-delimited components was
incorrect when the / character was escaped (i.e. \/), and a final \ at
the end of pattern was not handled correctly.
a v6 socket will only be used if there is at least one v6 nameserver
address. if the kernel lacks v6 support, the code will fall back to
using a v4 socket and requests to v6 servers will silently fail. when
using a v6 socket, v4 addresses are converted to v4-mapped form and
setsockopt is used to ensure that the v6 socket can accept both v4 and
v6 traffic (this is on-by-default on Linux but the default is
configurable in /proc and so it needs to be set explicitly on the
socket level). this scheme avoids increasing resource usage during
lookups and allows the existing network io loop to be used without
modification.
previously, nameservers whose address family did not match the address
family of the first-listed nameserver were simply ignored. prior to
recent __ipparse fixes, they were not ignored but erroneously parsed.
the old value of 20 was reported by Laurent Bercot as being
insufficient for a reasonable real-world usage case. actual problem
was the internal buffer used by ttyname(), but the implementation of
ttyname uses TTY_NAME_MAX, and for consistency it's best to increase
both. the new value is aligned with glibc.
subsequent code assumes the address family requested is either
unspecified or one of IPv4/IPv6, and could malfunction if this
constraint is not met, so other address families should be explicitly
rejected.
on archs with excess precision, the floating point constant 1e40f may
be evaluated such that it does not actually produce an infinity.
1e5000f is sufficiently large to produce an infinity for all supported
floating point formats. note that this definition of INFINITY is only
used for old or non-GNUC compilers anyway; despite being a portable,
conforming definition, it leads to erroneous warnings on many
compilers and thus using the builtin is preferred.
these functions were spuriously failing in the case where the buffer
size was exactly the number of bytes/characters to be written,
including null termination. since these functions do not have defined
error conditions other than buffer size, a reasonable application may
fail to check the return value when the format string and buffer size
are known to be valid; such an application could then attempt to use a
non-terminated buffer.
in addition to fixing the bug, I have changed the error handling
behavior so that these functions always null-terminate the output
except in the case where the buffer size is zero, and so that they
always write as many characters as possible before failing, rather
than dropping whole fields that do not fit. this actually simplifies
the logic somewhat anyway.
unfortunately this eliminates the ability of the compiler to diagnose
some dangerous/incorrect usage, but POSIX requires (as an extension to
the C language, i.e. CX shaded) that NULL have type void *. plain C
allows it to be defined as any null pointer constant.
the definition 0L is preserved for C++ rather than reverting to plain
0 to avoid dangerous behavior in non-conforming programs which use
NULL as a variadic sentinel. (it's impossible to use (void *)0 for C++
since C++ lacks the proper implicit pointer conversions, and other
popular alternatives like the GCC __null extension seem non-conforming
to the standard's requirements.)
loop condition was incorrect and confusing and caused an infinite loop
when (broken) applications reaped the pid from a signal handler or
another thread before wordexp's call to waitpid could do so.
when WRDE_NOSPACE is returned, the we_wordv and we_wordc members must
be valid, because the interface contract allows them to return partial
results.
in the case of zero results (due either to resource exhaustion or a
zero-word input) the we_wordv array still should contain a terminating
null pointer and the initial we_offs null pointers. this is impossible
on resource exhaustion, so a correct application must presumably check
for a null pointer in we_wordv; POSIX however seems to ignore the
issue. the previous code may have crashed under this situation.
avoid using exit status to determine if a shell error occurred, since
broken programs may install SIGCHLD handlers which reap all zombies,
including ones that don't belong to them. using clone and __WCLONE
does not seem to work for avoiding this problem since exec resets the
exit signal to SIGCHLD.
instead, the new code uses a dummy word at the beginning of the
shell's output, which is ignored, to determine whether the command was
executed successfully. this also fixes a corner case where a word
string containing zero words was interpreted as a single zero-length
word rather than no words at all. POSIX does not seem to require this
case to be supported anyway, though.
in addition, the new code uses the correct retry idiom for waitpid to
ensure that spurious STOP/CONT signals in the child and/or EINTR in
the parent do not prevent successful wait for the child, and blocks
signals in the child.
previously these macros wrongly had type double rather than long
double. I see no way an application could detect the error in C99, but
C11's _Generic can trivially detect it.
at the same time, even though these archs do not have excess
precision, the number of decimal places used to represent these
constants has been increased to 21 to be consistent with the decimal
representations used for the DBL_* macros.
this is enough to produce the correct value even if the constant is
interpreted as 80-bit extended precision, which matters on archs with
excess precision (FLT_EVAL_METHOD==2) under at least some
interpretations of the C standard. the shorter representations, while
correct if converted to the nominal precision at translation time,
could produce an incorrect value at extended precision, yielding
results such as (double)DBL_MAX != DBL_MAX.
this should not matter since the reality is that either all the sysv
sem syscalls are individual syscalls, or all of them are multiplexed
on the SYS_ipc syscall (depending on arch). but best to be consistent
anyway.
siginfo_t is not available from signal.h when the strict ISO C feature
profile (e.g. passing -std=c99 to gcc without defining any other
feature test macros) is used, but the type is needed to declare
waitid. using sys/wait.h (or any POSIX headers) in strict ISO C mode
is an application bug, but in the interest of compatibility, it's best
to avoid producing gratuitous errors. the simplest fix I could find is
suppressing the declaration of waitid (and also signal.h inclusion,
since it's not needed for anything else) in this case, while still
exposing everything else in sys/wait.h
it's not clear why I originally wrote O_NOFOLLOW into this; I suspect
the reason was with an aim of making the function more general for
mapping partially or fully untrusted files provided by the user.
however, the timezone code already precludes use of absolute or
relative pathnames in suid/sgid programs, and disallows .. in
pathnames which are relative to one of the system timezone locations,
so there is no threat of opening a symlink which is not trusted by
appropriate user. since some users may wish to put symbolic links in
the zoneinfo directories to alias timezones, it seems preferable to
allow this.
the rest of the code is not prepared to handle an empty TZ string, so
falling back to __gmt ("GMT"), just as if TZ had been blank or unset,
is the preferable action.
inet_aton returns a boolean success value, whereas __ipparse returns 0
on success and -1 on failure. also change the conditional in inet_addr
to be consistent with other uses of __ipparse where only negative
values are treated as failure.
now that we're waiting for the exit status of the child process, the
result can be conveyed in the exit status rather than via a pipe.
since the error value might not fit in 7 bits, a table is used to
translate possible meaningful error values to small integers.
I mistakenly assumed that clone without a signal produced processes
that would not become zombies; however, waitpid with __WCLONE is
required to release their pids.
while using "l" unconditionally gave the right behavior due to
matching sizes/representations, it was technically UB and produced
compiler warnings with format string checking.
despite being marked legacy, this was specified by SUSv3 as part of
the XSI option; only the most recent version of the standard dropped
it. reportedly there's actual code using it.
despite being practically deprecated, these functions are still part
of the standard and thus cannot reside in a file that also contains
namespace pollution. this reverts some of the changes made in commit
e40f48a421.
fcntl.h: AT_* is not a reserved namespace so extensions cannot be
exposed by default.
langinfo.h: YESSTR and NOSTR were removed from the standard.
limits.h: NL_NMAX was removed from the standard.
signal.h: the conditional for NSIG was wrongly checking _XOPEN_SOURCE
rather than _BSD_SOURCE. this was purely a mistake; it doesn't even
match the commit message from the commit that added it.
this fixes an issue reported by Daniel Thau whereby faccessat with the
AT_EACCESS flag did not work in cases where the process is running
suid or sgid but without root privileges. per POSIX, when the process
does not have "appropriate privileges", setuid changes the euid, not
the real uid, and the target uid must be equal to the current real or
saved uid; if this condition is not met, EPERM results. this caused
the faccessat child process to fail.
using the setreuid syscall rather than setuid works. POSIX leaves it
unspecified whether setreuid can set the real user id to the effective
user id on processes without "appropriate privileges", but Linux
allows this; if it's not allowed, there would be no way for this
function to work.
based on patch by Michael Forney. at the same time, I've changed the
if branch to be more clear, avoiding the comma operator.
the underlying issue is that Linux always returns ERANGE when size is
too short, even when it's zero, rather than returning EINVAL for the
special case of zero as required by POSIX.
there is no reason to check the return value for setting errno, since
brk never returns errors, only the new value of the brk (which may be
the same as the old, or otherwise differ from the requested brk, on
failure).
it may be beneficial to eventually just eliminate this file and make
the syscalls inline in malloc.c.
the va_arg call for the argv[]-terminating null pointer was missing,
so this pointer was being wrongly used as the environment pointer.
issue reported by Timo Teräs. proposed patch slightly modified to
simplify the resulting code.
bug report and patch by Michael Forney. the terminating null pointer
at the end of the gr_mem array was overwriting the beginning of the
string data, causing the gr_name member to always be a zero-length
string.
issue reported by Michael Forney:
"If wn becomes 0 after processing a chunk of 4, mbsrtowcs currently
continues on, wrapping wn around to -1, causing the rest of the string
to be processed.
This resulted in buffer overruns if there was only space in ws for wn
wide characters."
the original patch submitted added an additional check for !wn after
the loop; to avoid extra branching, I instead just changed the wn>=4
check to wn>=5 to ensure that at least one slot remains after the
word-at-a-time loop runs. this should not slow down the tail
processing on real-world usage, since an extra slot that can't be
processed in the word-at-a-time loop is needed for the null
termination anyway.
atomic store was lacking a barrier, which was fine for legacy arm with
no real smp and kernel-emulated cas, but unsuitable for more modern
systems. the kernel provides another "kuser" function, at 0xffff0fa0,
which could be used for the barrier, but using that would drop support
for kernels 2.6.12 through 2.6.14 unless an extra conditional were
added to check for barrier availability. just using the barrier in the
kernel cas is easier, and, based on my reading of the assembly code in
the kernel, does not appear to be significantly slower.
at the same time, other atomic operations are adapted to call the
kernel cas function directly rather than using a_cas; due to small
differences in their interface contracts, this makes the generated
code much simpler.
if a multithreaded program became non-multithreaded (i.e. all other
threads exited) while one thread held an internal lock, the remaining
thread would fail to release the lock. the the program then became
multithreaded again at a later time, any further attempts to obtain
the lock would deadlock permanently.
the underlying cause is that the value of libc.threads_minus_1 at
unlock time might not match the value at lock time. one solution would
be returning a flag to the caller indicating whether the lock was
taken and needs to be unlocked, but there is a simpler solution: using
the lock itself as such a flag.
note that this flag is not needed anyway for correctness; if the lock
is not held, the unlock code is harmless. however, the memory
synchronization properties associated with a_store are costly on some
archs, so it's best to avoid executing the unlock code when it is
unnecessary.
this was resulting in crashes in posix_spawn on mips, and would have
affected applications calling clone too. since the prototype for
__clone has it as a variadic function, it may not assume that 16($sp)
is writable for use in making the syscall. instead, it needs to
allocate additional stack space, and then adjust the stack pointer
back in both of the code paths for the parent process/thread.
CLONE_PARENT is not necessary (CLONE_THREAD provides all the useful
parts of it) and Linux treats CLONE_PARENT as an error in certain
situations, without noticing that it would be a no-op due to
CLONE_THREAD. this error case prevents, for example, use of a
multi-threaded init process and certain usages with containers.
unlike other archs, the mips version of clone was not doing anything
to align the stack pointer. this seems to have been the cause for some
SIGBUS crashes that were observed in posix_spawn.
msg.h was wrong for big-endian (wrong endiannness padding).
shm.h was just plain wrong (mips is not supposed to have padding).
both changes were tested using libc-test on qemu-system-mips.
the underlying problem was not incorrect sign extension (fixed in the
previous commit to this file by nsz) but that code that treats "long"
as 32-bit was copied blindly from i386 to x86_64.
now lrintl is identical to llrintl on x86_64, as it should be.
if fopen fails for a reason other than ENOENT, we must assume the
intent is that the path file be used. failure may be due to
misconfiguration or intentional resource-exhaustion attack (against
suid programs), in which case falling back to loading libraries from
an unintended path could be dangerous.
this protects against deadlock from spurious signals (e.g. sent by
another process) arriving after the controlling thread releases the
other threads from the sync operation.
the head pointer was not being reset between calls to synccall, so any
use of this interface more than once would build the linked list
incorrectly, keeping the (now invalid) list nodes from the previous
call.
invalid format strings invoke undefined behavior, so this is not a
conformance issue, but it's nicer for scanf to report the error safely
instead of calling free on a potentially-uninitialized pointer or a
pointer to memory belonging to the caller.
rather than allocating a PATH_MAX-sized buffer when the caller does
not provide an output buffer, work first with a PATH_MAX-sized temp
buffer with automatic storage, and either copy it to the caller's
buffer or strdup it on success. this not only avoids massive memory
waste, but also avoids pulling in free (and thus the full malloc
implementation) unnecessarily in static programs.
this avoids failure if the file is not readable and avoids odd
behavior for device nodes, etc. on old kernels that lack O_PATH, the
old behavior (O_RDONLY) will naturally happen as the fallback.
commit 07827d1a82 seems to have
introduced this issue. sigqueue is called from the synccall core, at
which time, even implementation-internal signals are blocked. however,
pthread_sigmask removes the implementation-internal signals from the
old mask before returning, so that a process which began life with
them blocked will not be able to save a signal mask that has them
blocked, possibly causing them to become re-blocked later. however,
this was causing sigqueue to unblock the implementation-internal
signals during synccall, leading to deadlock.
the BSD and GNU versions of this structure differ, so exposing it in
the default _BSD_SOURCE profile is possibly problematic. both versions
could be simultaneously supported with anonymous unions if needed in
the future, but for now, just omitting it except under _GNU_SOURCE
should be safe.
I originally added this warning option based on a misunderstanding of
how it works. it does not warn whenever the destination of the cast
has stricter alignment; it only warns in cases where misaligned
dereference could lead to a fault. thus, it's essentially a no-op for
i386, which had me wrongly believing the code was clean for this
warning level. on other archs, numerous diagnostic messages are
produced, and all of them are false-positives, so it's better just not
to use it.
unlike the old C memcpy, this version handles word-at-a-time reads and
writes even for misaligned copies. it does not require that the cpu
support misaligned accesses; instead, it performs bit shifts to
realign the bytes for the destination.
essentially, this is the C version of the ARM assembly language
memcpy. the ideas are all the same, and it should perform well on any
arch with a decent number of general-purpose registers that has a
barrel shift operation. since the barrel shifter is an optional cpu
feature on microblaze, it may be desirable to provide an alternate asm
implementation on microblaze, but otherwise the C code provides a
competitive implementation for "generic risc-y" cpu archs that should
alleviate the urgent need for arch-specific memcpy asm.
while the incorporation of this requirement from C99 into C++11 was
likely an accident, some software expects it to be defined, and it
doesn't hurt. if the requirement is removed, then presumably
__bool_true_false_are_defined would just be in the implementation
namespace and thus defining it would still be legal.
this version of memset is optimized both for small and large values of
n, and makes no misaligned writes, so it is usable (and near-optimal)
on all archs. it is capable of filling up to 52 or 56 bytes without
entering a loop and with at most 7 branches, all of which can be fully
predicted if memset is called multiple times with the same size.
it also uses the attribute extension to inform the compiler that it is
violating the aliasing rules, unlike the previous code which simply
assumed it was safe to violate the aliasing rules since translation
unit boundaries hide the violations from the compiler. for non-GNUC
compilers, 100% portable fallback code in the form of a naive loop is
provided. I intend to eventually apply this approach to all of the
string/memory functions which are doing word-at-a-time accesses.
this will be needed for upcoming commits to the string/mem functions
to correct their unannounced use of aliasing violations for
word-at-a-time search, fill, and copy operations.
this is a nonstandard extension but will be required in the next
version of POSIX, and it's widely used/useful in shell scripts
utilizing the date utility.
Rich Felker