Signed-off-by: Simon McVittie <smcv@collabora.com>

This means new symbols actually get picked up.

Signed-off-by: Simon McVittie <smcv@collabora.com>

Signed-off-by: Simon McVittie <smcv@collabora.com>

The dlsym wrapper needs to cache the implementation of dlsym that's
in the GOT at the point when it (the wrapper) is installed.

Moving the implementation of the wrapper helper to capsule_init's
file so that capsule_init can set some static state to do this for
us.

This is enough to make openarena work.

If someone actually reaches a shim stubroutine the stack
may be corrupt - the stubroutines don't have the same signature
as the functions they act as placeholders for.

Rather than risk proceeding with a borked stack we now abort once
we've tried to dump a backtrace.

warning: ordered comparison of pointer with integer zero [-Wextra]

Signed-off-by: Simon McVittie <smcv@collabora.com>

Signed-off-by: Simon McVittie <smcv@collabora.com>

Signed-off-by: Simon McVittie <smcv@collabora.com>

This could be a bug if the implicit declaration doesn't have the
same ABI that we actually implement.

Signed-off-by: Simon McVittie <smcv@collabora.com>

Signed-off-by: Simon McVittie <smcv@collabora.com>

This silences -Wstrict-prototypes.

Signed-off-by: Simon McVittie <smcv@collabora.com>

Signed-off-by: Simon McVittie <smcv@collabora.com>

Signed-off-by: Simon McVittie <smcv@collabora.com>

Signed-off-by: Simon McVittie <smcv@collabora.com>

This should be enabled in basically all C code.

Signed-off-by: Simon McVittie <smcv@collabora.com>

Signed-off-by: Simon McVittie <smcv@collabora.com>

This means we can be overridden with LIBELF_CFLAGS, LIBELF_LIBS in the
standard way.

Signed-off-by: Simon McVittie <smcv@collabora.com>

Signed-off-by: Simon McVittie <smcv@collabora.com>

Signed-off-by: Simon McVittie <smcv@collabora.com>

Signed-off-by: Simon McVittie <smcv@collabora.com>

Signed-off-by: Simon McVittie <smcv@collabora.com>

Debian Policy §8 says so.

Signed-off-by: Simon McVittie <smcv@collabora.com>

bash is Essential: yes, so Policy says we should not depend on it.

Signed-off-by: Simon McVittie <smcv@collabora.com>

Signed-off-by: Simon McVittie <smcv@collabora.com>

This affects the order of the linker command line in ways that matter
if you have a linker that defaults to --as-needed, like Ubuntu's, or
if you are linking statically. Dependency libraries (LDADD/LIBADD)
must appear last on the link line (so that they can satisfy
unresolved symbols of earlier items on the link line), while LDFLAGS
appear first.

Signed-off-by: Simon McVittie <smcv@collabora.com>

Since it's completely generic, we should let autogen.sh create it.

Signed-off-by: Simon McVittie <smcv@collabora.com>

R_X86_64_GLOB_DAT and R_X86_64_64 are now supported.

The capsule-init-project script now implements and installs a
dlsym() wrapper in generated shim project source trees.

In addition to wrapping dlopen() inside the capsule we need to
wrap dlsym() outside the capsule in order to allow libraries
that use a dlopen()/dlsym() pattern to work transparently.

capsule_shim_dlsym() bundles most of the standard logic we
expect such wrappers to use.

We shouldn't ever reach a shim stubroutine, so if we do it means
a relocation has probably failed: We dump a few backtrace frames
if this happens so we can track down which DSO called into the
shim, and hopefully figure out what kind of relocation it was.

Specifically: Can't handle libelf ElfW(X) style type declarations
inside a struct (but can if it's a free-standing typedef).