On Mon, Feb 10, 2020 at 8:22 PM H.J. Lu <[email protected]> wrote:
>
> Since nested function isn't only called directly, there is ENDBR32 at
> function entry and we need to skip it for direct jump in trampoline.
Hm, I'm afraid I don't understand this comment. Can you perhaps rephrase it?
Uros.
> Tested on Linux/x86-64 CET machine with and without -m32.
>
> gcc/
>
> PR target/93656
> * config/i386/i386.c (ix86_trampoline_init): Skip ENDBR32 at
> nested function entry.
>
> gcc/testsuite/
>
> PR target/93656
> * gcc.target/i386/pr93656.c: New test.
> ---
> gcc/config/i386/i386.c | 7 ++++++-
> gcc/testsuite/gcc.target/i386/pr93656.c | 4 ++++
> 2 files changed, 10 insertions(+), 1 deletion(-)
> create mode 100644 gcc/testsuite/gcc.target/i386/pr93656.c
>
> diff --git a/gcc/config/i386/i386.c b/gcc/config/i386/i386.c
> index 44bc0e0176a..dbcae244acb 100644
> --- a/gcc/config/i386/i386.c
> +++ b/gcc/config/i386/i386.c
> @@ -16839,9 +16839,14 @@ ix86_trampoline_init (rtx m_tramp, tree fndecl, rtx
> chain_value)
> the stack, we need to skip the first insn which pushes the
> (call-saved) register static chain; this push is 1 byte. */
> offset += 5;
> + int skip = MEM_P (chain) ? 1 : 0;
> + /* Since nested function isn't only called directly, there is
> + ENDBR32 at function entry and we need to skip it. */
> + if (need_endbr)
> + skip += 4;
> disp = expand_binop (SImode, sub_optab, fnaddr,
> plus_constant (Pmode, XEXP (m_tramp, 0),
> - offset - (MEM_P (chain) ? 1 : 0)),
> + offset - skip),
> NULL_RTX, 1, OPTAB_DIRECT);
> emit_move_insn (mem, disp);
> }
> diff --git a/gcc/testsuite/gcc.target/i386/pr93656.c
> b/gcc/testsuite/gcc.target/i386/pr93656.c
> new file mode 100644
> index 00000000000..f0ac8c8edaa
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/i386/pr93656.c
> @@ -0,0 +1,4 @@
> +/* { dg-do run { target { ia32 && cet } } } */
> +/* { dg-options "-O2 -fcf-protection" } */
> +
> +#include "pr67770.c"
> --
> 2.24.1
>
