[sb2] Fix can_accelerate_rpm under new lua. Fixes JB#54116

See merge request !52

The (first) return value of os.execute has changed between lua 5.1
and 5.2. On old lua, successful execution returned 0, failure
(exit 1 in script) returned 256. On new lua, successful execution
returns true and failure false.

This reverts commit 3fb3ee27.

[sb2] Fix x86 sem_open error. Contributes to jb#53896

See merge request !51

Support /usr merge by allowing /bin/perl to be accelerated JB#36606

See merge request !23

Now that /bin is a symlink to /usr/bin it seems that sometimes perl is
found as /bin/perl. This can manifest in things like autoreconf having
a shebang of #!/bin/perl (Speculatively because of PATH having /bin
first?)

This misses the /usr/bin/perl rules in a way which means that the ARM
perl tries to load x86 .so extensions.
Signed-off-by: David Greaves <david.greaves@jolla.com>

nomap_nolog variants of fstat64 and stat64 are used when `_STAT_VER`
is not defined (glibc >=2.33) but were not generated.

`table.maxn()` is deprecated with lua 5.2 and removed by lua 5.4.

Instead of using `lua_call` use `lua_pcall` to exit gracefully in case we
hit any lua errors.

[preload] Handle O_NOFOLLOW in vperm if flags are used.

See merge request !46

Disable llvm compiler acceleration.

See merge request !48

Signed-off-by: Matti Kosola <matti.kosola@jolla.com>

Accelerate rust and cargo. Contributes JB#50499

See merge request !47

Signed-off-by: David Greaves <david.greaves@jolla.com>

Fix "Unrecognized escape \M passed through" warning.

[sb2] Fix typo. JB#52244

See merge request !45

[sb2] Fix make depency graph resolving. Fixes JB#52244

See merge request !44

The output directory has to exist before make creates the dependency
graph or the build may break when CWD is not $(SCRDIR).
This for example caused in case of rule_tree.o make to have to receipes for the same file one with -fPIC
and one without -fPIC breaking PIC and breaking building libsb2.so on x86_64.

Fix host-gcc use

See merge request !41

It has been obsoleted since commit
1b7fb2f6 (Create /usr/bin "autorules"
via init.lua) and commit c85c5397
(Top-level Makefile: Install rule_constants.lua) stopped installing it.

This fixes commit dbcf9d60 (support
different gcc search patchs)

Initially, my_envp equals to *new_envp, then it is modified. Those
modifications are lost when *new_envp is used after that.

The (possibly) discarted modifications to env. variables affect
SBOX_DISABLE_MAPPING, SBOX_DISABLE_ARGVENVP and __SB2_REAL_BINARYNAME.

Regarding SBOX_DISABLE_MAPPING, consider a host executable being
executed by sb2 and trying to execute another host executable. With
mapping not disabled, the other executable may be mapped to a different
executable under target/tools root instead. In most cases this is likely
to happen unnoticed. It breaks when there is an incompatibility between
those, e.g., when the host executable is a GCC compiler trying to
execute some of its internal tools - an i486 host compiler may end up
executing an arm assembler this way.

Trying to explain why *new_envp was used here, I don't believe it was a
typo. At the same time I don't believe it was intentional to change the
logic this way. I can't think of a reason to discard changes to the
afforementioned env. vars. To me it seems inevitable to keep them set.
Also, change was introduced with commit
3f042f36 (Execs: Postprocess native
dynamic binaries with C code) whose only objective seems to be
reimplementation of some Lua code in C.  Maybe it was done to deal with
some compiler bug and no one noticed it breaks things, as this only
breaks in rather special cases like the one mentioned above.

[sb2] Add support for posix_spawnp(). Contributes to JB#51819

Move it into a shared function and reuse errno from access(2).