On 01/14/15 04:32, Pratyush Anand wrote:

On Sun, Jan 11, 2015 at 9:33 AM, David Long [off-list ref] wrote:

quoted

From: "David A. Long" <redacted>

Certain instructions are hard to execute correctly out-of-line (as in
kprobes).  Test functions are added to insn.[hc] to identify these.  The
instructions include any that use PC-relative addressing, change the PC,
or change interrupt masking. For efficiency and simplicity test
functions are also added for small collections of related instructions.

Signed-off-by: David A. Long <redacted>
---
  arch/arm64/include/asm/insn.h | 21 +++++++++++++++++++--
  arch/arm64/kernel/insn.c      | 18 ++++++++++++++++++
  2 files changed, 37 insertions(+), 2 deletions(-)

diff --git a/arch/arm64/include/asm/insn.h b/arch/arm64/include/asm/insn.h
index e2ff32a..466afd4 100644
--- a/arch/arm64/include/asm/insn.h
+++ b/arch/arm64/include/asm/insn.h

@@ -223,8 +223,13 @@ static __always_inline bool aarch64_insn_is_##abbr(u32 code) \
  static __always_inline u32 aarch64_insn_get_##abbr##_value(void) \
  { return (val); }

+__AARCH64_INSN_FUNCS(adr,      0x9F000000, 0x10000000)

Should n't it be
__AARCH64_INSN_FUNCS(adr_adrp,      0x1F000000, 0x10000000)

So, that it also take care about adrp

Yes, that does look like a mistake.

quoted

+__AARCH64_INSN_FUNCS(prfm_lit, 0xFF000000, 0xD8000000)

[...]

quoted

+bool aarch64_insn_uses_literal(u32 insn)
+{
+       /* ldr/ldrsw (literal), prfm */
+
+       return aarch64_insn_is_ldr_lit(insn) ||
+               aarch64_insn_is_ldrsw_lit(insn) ||

also aarch64_insn_is_adr_adrp(insn) ||

Yup.

quoted

+               aarch64_insn_is_prfm_lit(insn);
+}
+
+bool aarch64_insn_is_branch(u32 insn)
+{
+       /* b, bl, cb*, tb*, b.cond, br, blr */
+
+       return aarch64_insn_is_b_bl_cb_tb(insn) ||
+               aarch64_insn_is_br_blr(insn) ||

also aarch64_insn_is_ret(insn) ||

The goal was to catch intructions that use a PC-relative branch, since 
the PC will not be what is expected.  Of course any instruction that 
changes the PC will have a problem too because the PC will be rewritten 
after the probe is completed.  So, yeah, this needs to be fixed.

quoted

+               aarch64_insn_is_bcond(insn);
+}
+
  /*

-dl