add -fret-clean option (amd64 and i386 only at first), defaulting to off.
This causes the caller to cleans the return address off the stack after
a callq completes. The option is best used in low-le

add -fret-clean option (amd64 and i386 only at first), defaulting to off.
This causes the caller to cleans the return address off the stack after
a callq completes. The option is best used in low-level libraries (such as
libc), because libc contains low-level system call stubs. The option
reduces hints (found on the stale parts of the stack) about libc.so's mapping
location, and together with random-relinking, relro got/pic, and xonly
makes some exploit methods more difficult.
ok mortimer, mlarkin, much discussion with kettenis, in snaps for 2 weeks.

show more ...

merge clang-16.0.6

Merge LLVM 13.0.0.

Merge LLVM 11.1.0.

switch clang to -fno-common by default

Common variables will be flagged as errors by default. This can
be overridden with -fcommon.

Cherry-pick this change that will be part of LLVM11. We bring i

switch clang to -fno-common by default

Common variables will be flagged as errors by default. This can
be overridden with -fcommon.

Cherry-pick this change that will be part of LLVM11. We bring it
forward so we can finish fixing the fallout in the ports tree.

ok deraadt@ sthen@ mglocker@

show more ...

Implement Loongson 2F BTB bug workaround in clang

This adapts the -mfix-loongson2f-btb workaround from as(1) to
LLVM/clang. Because LLVM emits machine code directly without using
assembly as interme

Implement Loongson 2F BTB bug workaround in clang

This adapts the -mfix-loongson2f-btb workaround from as(1) to
LLVM/clang. Because LLVM emits machine code directly without using
assembly as intermediary form, the workaround has to be applied
in the compiler backend in addition to the integrated assembler.

Discussed with miod@
OK mortimer@

show more ...

Disable strict floating point if not X86

Disable options like clang -fno-unsafe-math-optimizations for all
targets except X86 (and SystemZ). This prevents a crash of clang,
reported by cwen@, in gr

Disable strict floating point if not X86

Disable options like clang -fno-unsafe-math-optimizations for all
targets except X86 (and SystemZ). This prevents a crash of clang,
reported by cwen@, in graphics/babl or emulators/mednafen for powerpc.

ok mortimer@

This change comes from llvm 11.x:
https://github.com/llvm/llvm-project/commit/d4ce862

Reland "[FPEnv][Clang][Driver] Disable constrained floating point on targets lacking support."

We currently have strict floating point/constrained floating point enabled
for all targets. Constrained SDAG nodes get converted to the regular ones
before reaching the target layer. In theory this should be fine.

However, the changes are exposed to users through multiple clang options
already in use in the field, and the changes are _completely_ _untested_
on almost all of our targets. Bugs have already been found, like
"https://bugs.llvm.org/show_bug.cgi?id=45274".

This patch disables constrained floating point options in clang everywhere
except X86 and SystemZ. A warning will be printed when this happens.

Use the new -fexperimental-strict-floating-point flag to force allowing
strict floating point on hosts that aren't already marked as supporting
it (X86 and SystemZ).

Differential Revision: https://reviews.llvm.org/D80952

show more ...

Import LLVM 10.0.1 including clang, lld and lldb.

Import LLVM 10.0.0 release including clang, lld and lldb.

ok hackroom
tested by plenty