It has been a long practice for software using IEEE 754 floating-point
arithmetic run on MIPS processors to use an encoding of Not-a-Number
(NaN) data different to one used by software run on other processors.
And as of IEEE 754-2008 revision [1] this encoding does not follow one
recommended in the standard, as specified in section 6.2.1, where it
is stated that quiet NaNs should have the first bit (d1) of their
significand set to 1 while signalling NaNs should have that bit set to
0, but MIPS software interprets the two bits in the opposite manner.
As from revision 3.50 [2][3] the MIPS Architecture provides for
processors that support the IEEE 754-2008 preferred NaN encoding format.
As the two formats (further referred to as "legacy NaN" and "2008 NaN")
are incompatible to each other, tools have to provide support for the
two formats to help people avoid using incompatible binary modules.
The change is comprised of two functional groups of features, both of
which are required for correct support.
1. Dynamic linker support.
To enforce the NaN encoding requirement in dynamic linking a new ELF
file header flag has been defined. This flag is set for 2008-NaN
shared modules and executables and clear for legacy-NaN ones. The
dynamic linker silently ignores any incompatible modules it
encounters in dependency processing.
To avoid unnecessary processing of incompatible modules in the
presence of a shared module cache, a set of new cache flags has been
defined to mark 2008-NaN modules for the three ABIs supported.
Changes to sysdeps/unix/sysv/linux/mips/readelflib.c have been made
following an earlier code quality suggestion made here:
http://sourceware.org/ml/libc-ports/2009-03/msg00036.html
and are therefore a little bit more extensive than the minimum
required.
Finally a new name has been defined for the dynamic linker so that
2008-NaN and legacy-NaN binaries can coexist on a single system that
supports dual-mode operation and that a legacy dynamic linker that
does not support verifying the 2008-NaN ELF file header flag is not
chosen to interpret a 2008-NaN binary by accident.
2. Floating environment support.
IEEE 754-2008 features are controlled in the Floating-Point Control
and Status (FCSR) register and updates are needed to floating
environment support so that the 2008-NaN flag is set correctly and
the kernel default, inferred from the 2008-NaN ELF file header flag
at the time an executable is loaded, respected.
As the NaN encoding format is a property of GCC code generation that is
both a user-selected GCC configuration default and can be overridden
with GCC options, code that needs to know what NaN encoding standard it
has been configured for checks for the __mips_nan2008 macro that is
defined internally by GCC whenever the 2008-NaN mode has been selected.
This mode is determined at the glibc configuration time and therefore a
few consistency checks have been added to catch cases where compilation
flags have been overridden by the user.
The 2008 NaN set of features relies on kernel support as the in-kernel
floating-point emulator needs to be aware of the NaN encoding used even
on hard-float processors and configure the FPU context according to the
value of the 2008 NaN ELF file header flag of the executable being
started. As at this time work on kernel support is still in progress
and the relevant changes have not made their way yet to linux.org master
repository.
Therefore the minimum version supported has been artificially set to
10.0.0 so that 2008-NaN code is not accidentally run on a Linux kernel
that does not suppport it. It is anticipated that the version is
adjusted later on to the actual initial linux.org kernel version to
support this feature. Legacy NaN encoding support is unaffected, older
kernel versions remain supported.
[1] "IEEE Standard for Floating-Point Arithmetic", IEEE Computer
Society, IEEE Std 754-2008, 29 August 2008
[2] "MIPS Architecture For Programmers, Volume I-A: Introduction to the
MIPS32 Architecture", MIPS Technologies, Inc., Document Number:
MD00082, Revision 3.50, September 20, 2012
[3] "MIPS Architecture For Programmers, Volume I-A: Introduction to the
MIPS64 Architecture", MIPS Technologies, Inc., Document Number:
MD00083, Revision 3.50, September 20, 2012
Long ago static startup did not parse the auxiliary vector and therefore
could not get at any `AT_FPUCW' tag to check whether upon FPU context
allocation the kernel would use a FPU control word setting different to
that provided by the `__fpu_control' variable. Static startup therefore
always initialized the FPU control word, forcing immediate FPU context
allocation even for binaries that otherwise never used the FPU.
As from GIT commit f8f900ecb9 static
startup supports parsing the auxiliary vector, so now it can avoid
explicit initialization of the FPU control word, just as can dynamic
startup, in the usual case where the setting written to the FPU control
word would be the same as the kernel uses. This defers FPU context
allocation until the binary itself actually pokes at the FPU.
Note that the `AT_FPUCW' tag is usually absent from the auxiliary vector
in which case _FPU_DEFAULT is assumed to be the kernel default.
This change creates a link map in static executables to serve as the
global search list for dlopen. It fixes a problem with the inability
to access the global symbol object and a crash on an attempt to map a
DSO into the global scope. Some code that has become dead after the
addition of this link map is removed too and test cases are provided.
Resolves: #15465
The program name may be unavailable if the user application tampers
with argc and argv[]. Some parts of the dynamic linker caters for
this while others don't, so this patch consolidates the check and
fallback into a single macro and updates all users.
In dl-hwcaps.c the comment read that rounding was done
to ElfW(Addr), but it's actually rounded to ElfW(Word).
In ldconfig.c we make each comment a sentence and
mention that the "tls" pseudo-hwcap is just for legacy
installations where TLS was optional.
---
2013-05-22 Carlos O'Donell <carlos@redhat.com>
* elf/ldconfig.c (is_hwcap_platform): Make comments full setences.
(main): Mention "tls" pseudo-hwcap is legacy.
* elf/dl-hwcaps.c (_dl_important_hwcaps): Correct rounding comment.
Loading of the vDSO pseudo-hwcap from the type 2 GNU note is
a rather arcane and poorly documented process. Given that I had
a chance to review this code today I thought I would add all
of the things I had to lookup to verify the validity of the
process.
With a single .note.GNU the vDSO can register up to 64 flags,
though in practice you are limited to 64 - _DL_FIRST_EXTRA
bits which on x86 is 12 bits.
The only use of this that I know of is in the Xen support
in Linux where they use the 1st bit to indicate "nosegneg".
I see "We use bit 1 to avoid bugs in some versions of glibc
when bit 0 is used; the choice is otherwise arbitrary.", but
no reference to a glibc bug anywhere. The code as-is should
support bit zero, so we still have that free for future use.
The kernel, glibc, and ld.so.cache must coordinate to ensure
that bit values don't go too high and are used consistently.
---
2013-05-13 Carlos O'Donell <carlos@redhat.com>
* elf/dl-hwcaps.c (_dl_important_hwcaps): Comment vDSO hwcap loading.
* elf/ldconfig.c (is_hwcap_platform): Comment each hwcap check.
(main): Comment "tls" pseudo-hwcap.
The algorithm for scanning dependencies upon dlclose is
less than immediately obvious. This patch adds two bits
of comments that explain why you start the dependency
search at l_initfini[1], and why you need to restart
the search.
---
2013-05-09 Carlos O'Donell <carlos@redhat.com>
* elf/dl-close.c (_dl_close_worker): Add comments.
The seen array was doubled in size recently, but the memset to clear
the array was not adjusted. We adjust the memset to always be correct
regardless of the size of seen.
---
2013-04-06 Carlos O'Donell <carlos@redhat.com>
[BZ #15309]
* elf/dl-open.c (dl_open_worker): memset all of seen array.
These prototypes are duplicated in many places. Add a dedicated
header for holding prototypes for program-specific functions to
avoid that.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
ARM now supports loading unmarked objects from
the dynamic loader cache. Unmarked objects can
be used with the hard-float or soft-float ABI.
We must support loading unmarked objects during
the transition period from a binutils that does
not mark objects to one that does mark them with
the correct ELF flags.
Signed-off-by: Carlos O'Donell <carlos@redhat.com>
We setup $(READELF) and use it everywhere, so fix the two places
that were using readelf directly.
Reported-by: Denis M. <god@politeia.in>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* sysdeps/generic/ldconfig.h (FLAG_AARCH64_LIB64): New macro.
* elf/cache.c (print_entry): Print ",AArch64" for
FLAG_AARCH64_LIB64.
Signed-off-by: Steve McIntyre <steve.mcintyre@linaro.org>
Reviewed-by: Carlos O'Donell <carlos@systemhalted.org>
* sysdeps/generic/ldconfig.h (FLAG_ARM_LIBHF): New macro.
* elf/cache.c (print_entry): Print ",hard-float" for
FLAG_ARM_LIBHF.
Signed-off-by: Steve McIntyre <steve.mcintyre@linaro.org>
Reviewed-by: Carlos O'Donell <carlos@systemhalted.org>
This hook is useful for any arch-specific functionality that
should be done on loaded objects. For the tile architecture,
the hook is already provided (though we switch to using the new
macro name with this commit) and implements a simulator notifier
so that the simulator can load Elf symbols to match the object
and generate better error messages for PC's.
Also, remove a spurious definition of DL_UNMAP in dl-runtime.c
* elf/Makefile (tests): Remove conditional for have-initfini-array
since this is now always required and the variable does not exist
anymore.
(tests-static): Likewise.
(modules-names): Likewise.
* elf/tst-array1.c (fini_array): Make writeable so that it can be
merged with constructor/destructor.
(init_array): Likewise.
* elf/tst-array2dep.c (fini_array): Likewise.
(init_array): Likewise.
Commit glibc-2.14~10 disallowed rtld self loading to avoid a segfault
that used to happen when rtld was loading itself in normal mode.
Unfortunately, that commit disallowed all modes of self loading,
including those that used to work before. This change limits the check
for self loading to normal mode only, so that instruments like ldd could
handle rtld properly.
* csu/libc-tls.c (static_dtv): Renamed to ...
(_dl_static_dtv): This. Make it global.
(_dl_initial_dtv): Removed.
(__libc_setup_tls): Updated.
* elf/dl-tls.c (DL_INITIAL_DTV): New macro.
(_dl_deallocate_tls): Replace GL(dl_initial_dtv) with
DL_INITIAL_DTV.
When unmapping the first object in a namespace, the runtime linker
did not update the externally visible pointer. This resulted in
debuggers seeing pointers to memory that had been freed.
Add support for STT_GNU_IFUNC symbols and the new R_390_IRELATIVE
relocation. Provide optimized version of memcpy, memset, and memcmp
for z10 and z196.
[BZ #13579] Do not free l_initfini and allow it to be reused
on subsequent dl_open calls for the same library. This fixes
the invalid memory access in do_lookup_x when the previously
free'd l_initfini was accessed through l_searchlist when a
library had been opened for the second time.
[BZ #13882]
* elf/dl-deps.c (_dl_map_object_deps): Fix cycle detection. Use
uint16_t for elements in the "seen" array to avoid char overflows.
* elf/dl-fini.c (_dl_sort_fini): Likewise.
* elf/dl-open.c (dl_open_worker): Likewise.
When a stack is marked executable due to loading a DSO that requires
an executable stack, the logic tends to leave out a portion of stack
after the first frame, thus causing a difference in the value returned
by pthread_getattr_np before and after the stack is marked
executable. It ought to be possible to fix this by marking the rest of
the stack as executable too, but in the interest of marking as less of
the stack as executable as possible, the path this fix takes is to
make pthread_getattr_np also look at the first frame as the underflow
end of the stack and compute size and stack top accordingly.
The above happens only for the main process stack. NPTL thread stacks
are not affected by this change.
The current implementation of __strtoul_internal seems to only pretend
to support hex and octal strings by detecting a preceding 0x or 0 and
marking base as 8 or 16. When it comes to the actual processing of the
string, it only considers numeric values within, thus breaking hex
values that may have [a-f] in them. Fixed with this commit.
It may sometimes be desirable to make the dynamic linker only pick up
libraries from the library path and rpath and not look at the
ld.so.cache that ldconfig generates. An example of such a use case is
the glibc testsuite where the dynamic linker must not be influenced by
any external paths or caches.
This change adds a new option --inhibit-ldcache that when used, tells
the dynamic linker to not use ld.so.cache even if it is available.
* elf/rtld.c (dl_main): If DL_DEBUG_UNUSED is enabled, turn off
lazy binding.
* elf/dl-lookup (_dl_lookup_symbol_x): If DL_DEBUG_UNUSED, ignore
undefined symbol errors.
* elf/rtlc.c (dl_main): Skip VDSO when checking for unused
DT_NEEDED entries.
* sysdeps/sparc/sparc32/dl-machine.h (elf_machine_rela): Protect local
variables with appropriate CPP guards.
* sysdeps/sparc/sparc32/dl-trampoline.S: Propagate the stack_ptr from the
frame pointer, not the stack pointer. Correct layout comments. Fix test
on resulting framesize and the management of the outregs buffer for pltexit.
Preserve floating point return values across _dl_call_pltexit call.
* sysdeps/sparc/sparc64/dl-trampoline.S: Fix test on resulting
framesize and the management of the outregs buffer for pltexit.
Preserve floating point return values across _dl_call_pltexit
call.
* elf/sotruss-lib.c (la_sparc32_gnu_pltenter, la_sparc64_gnu_pltenter,
la_sparc32_gnu_pltexit, la_sparc64_gnu_pltexit): New functions.
(print_exit): Fix format string for return register value.
update_get_addr returns a link_map*, but the return expression
was of type char*. Fix by relying on gcc's void* arithmetic
extension and auto-conversion from void*.
We never seem to have hit this problem but way relative relocations
were skipped was wrong. There are relative relocations only in the
DT_REL/DT_RELA section. The elf_dynamic_do_##reloc function skipped
the entries in all calls, though.
When dynamically loading a library along with several dependencies, calls to
_dl_add_to_slotinfo and _dl_update_slotinfo can become intermixed. As a
consequence, _dl_update_slotinfo will update the generation counter of the dtv
although not all of the slots belonging to that generation have been added.
Subsequent calls to _dl_add_to_slotinfo will add more slots to the same
generation, for which no storage will be allocated, as the dtv generation
checks will claim no work is necessary. This will lead to uninitialized dtv
entries and will likely cause a SIGSEGV when thread local variables are
accessed.
The aux_cache fix of bug 11149 introduced a new bug - normally,
ldconfig -r never cares if the auxiliary cache is not available and
that is not a fatal problem, however this is not the case in case
of ldconfig -r when executed as non-root. In that case, ldconfig -r
fails hard unless var/cache/ldconfig/ exists within the chroot. This
patch fixes that.
Conflicts:
ChangeLog
GCC 4.5 warns about "extern void _end; &end;".
Use char[] instead, as that also doesn't fall foul
of a target's .sdata optimizations.
Signed-off-by: Richard Henderson <rth@twiddle.net>
If we should use further fields in the e_ident array in the ELF header
of files we want zero to mean the default. Enforce zero for now so
that non-optional features indicated by nonzero values are not go
unnoticed when using an old dynamic linker.
If a binary gets invoked by passing it as argument to ld.so the stack
still holds the auxiliary vector of ld.so when entering the _start
routine of the executable. So the invocation via ld.so is not fully
transparent to the executable. This causes problems if the executable
wants to scan the auxv itself.
On some architectures the update of the l_used field in the lookup
functions races with setting the other bits in the bitfield. Simply
avoid this and optimize use of l_used in general.
This patch fixes mixed SSE/AVX audit and checks AVX only once in
_dl_runtime_profile. When an AVX or SSE register value in pltenter is
modified, we have to make sure that the SSE part value is the same in both
lr_xmm and lr_vector fields so that pltexit will get the correct value
from either lr_xmm or lr_vector fields. AVX-enabled pltenter should
update both lr_xmm and lr_vector fields to support stacked AVX/SSE
pltenter functions.
SSE registers are used for passing parameters and must be preserved
in runtime relocations. This is inside ld.so enforced through the
tests in tst-xmmymm.sh. But the malloc routines used after startup
come from libc.so and can be arbitrarily complex. It's overkill
to save the SSE registers all the time because of that. These calls
are rare. Instead we save them on demand. The new infrastructure
put in place in this patch makes this possible and efficient.
There were several issues when the initial 31 entries hashtab filled up.
size * 3 <= tab->n_elements is always false, table can't have more elements
than its size. I assume from libiberty/hashtab.c this meant to be check for
3/4 full. Even after fixing that, _dl_higher_prime_number (31) apparently
returns 31, only _dl_higher_prime_number (32) returns 61. And, size
variable wasn't updated during reallocation, which means during reallocation
the insertion of the new entry was done into a wrong spot.
All this lead to a hang in ld.so, because a search with n_elements 31 size
31 wouldn't ever terminate.
