Re: linux-next: build warnings after merge of the powerpc-objtool tree

[Sathvika Vasireddy <sv@linux.ibm.com>]

Hi all,

On 25/11/22 09:00, Stephen Rothwell wrote:
> Hi all,
>
> After merging the powerpc-objtool tree, today's linux-next build (powerpc
> pseries_le_defconfig) produced these warnings:
>
> arch/powerpc/kernel/head_64.o: warning: objtool: end_first_256B(): can't find starting instruction
> arch/powerpc/kernel/optprobes_head.o: warning: objtool: optprobe_template_end(): can't find starting instruction
>
> I have no idea what started this (they may have been there yesterday).
I was able to recreate the above mentioned warnings with 
pseries_le_defconfig and powernv_defconfig. The regression report also 
mentions a warning 
(https://lore.kernel.org/oe-kbuild-all/202211282102.QUr7HHrW-lkp@intel.com/) 
seen with arch/powerpc/kernel/kvm_emul.S assembly file.

  [1] arch/powerpc/kernel/optprobes_head.o: warning: objtool: 
optprobe_template_end(): can't find starting instruction
  [2] arch/powerpc/kernel/kvm_emul.o: warning: objtool: 
kvm_template_end(): can't find starting instruction
  [3] arch/powerpc/kernel/head_64.o: warning: objtool: end_first_256B(): 
can't find starting instruction

The warnings [1] and [2] go away after adding 'nop' instruction. Below 
diff fixes it for me:

diff --git a/arch/powerpc/kernel/optprobes_head.S 
b/arch/powerpc/kernel/optprobes_head.S
index cd4e7bc32609..ea4e3bd82f4f 100644
--- a/arch/powerpc/kernel/optprobes_head.S
+++ b/arch/powerpc/kernel/optprobes_head.S
@@ -134,3 +134,4 @@ optprobe_template_ret:

         .global optprobe_template_end
  optprobe_template_end:
+       nop

diff --git a/arch/powerpc/kernel/kvm_emul.S b/arch/powerpc/kernel/kvm_emul.S
index 7af6f8b50c5d..41fd664e3ba0 100644
--- a/arch/powerpc/kernel/kvm_emul.S
+++ b/arch/powerpc/kernel/kvm_emul.S
@@ -352,3 +352,4 @@ kvm_tmp_end:

  .global kvm_template_end
  kvm_template_end:
+       nop

For warning [3], objtool is throwing can't find starting instruction 
warning because it finds that the symbol (end_first_256B) is zero sized, 
and such symbols are not added to the rbtree. I tried to fix it by 
adding a 'nop' instruction (pasted diff below), but that resulted in a 
kernel build failure.

diff --git a/arch/powerpc/kernel/head_64.S b/arch/powerpc/kernel/head_64.S
index 874efd25cc45..d48850fe159f 100644
--- a/arch/powerpc/kernel/head_64.S
+++ b/arch/powerpc/kernel/head_64.S
@@ -192,6 +192,7 @@ __secondary_hold:
         EMIT_BUG_ENTRY 0b, __FILE__, __LINE__, 0
  #endif
  CLOSE_FIXED_SECTION(first_256B)
+nop

  /*
   * On server, we include the exception vectors code here as it

diff --git a/arch/powerpc/kernel/exceptions-64s.S 
b/arch/powerpc/kernel/exceptions-64s.S
index 26f8fef53c72..f7517d443e9b 100644
--- a/arch/powerpc/kernel/exceptions-64s.S
+++ b/arch/powerpc/kernel/exceptions-64s.S
@@ -3104,9 +3104,13 @@ __end_interrupts:
  DEFINE_FIXED_SYMBOL(__end_interrupts, virt_trampolines)

  CLOSE_FIXED_SECTION(real_vectors);
+nop
  CLOSE_FIXED_SECTION(real_trampolines);
+nop
  CLOSE_FIXED_SECTION(virt_vectors);
+nop
  CLOSE_FIXED_SECTION(virt_trampolines);
+nop

  USE_TEXT_SECTION()

I'm not very sure on how to address this particular warning 
(arch/powerpc/kernel/head_64.o: warning: objtool: end_first_256B(): 
can't find starting instruction). Given that there are no calls to 
_mcount, one workaround is to skip objtool from running on 
arch/powerpc/kernel/head_64.o file. The below diff works for me:

diff --git a/arch/powerpc/kernel/Makefile b/arch/powerpc/kernel/Makefile
index 9b6146056e48..9ef6a040d875 100644
--- a/arch/powerpc/kernel/Makefile
+++ b/arch/powerpc/kernel/Makefile
@@ -219,3 +219,5 @@ $(obj)/vdso64_wrapper.o : $(obj)/vdso/vdso64.so.dbg

  # for cleaning
  subdir- += vdso
+
+OBJECT_FILES_NON_STANDARD_head_64.o := y


Thanks,
Sathvika