Re: [PATCHv3 0/8] x86/Linux Target Description Changes

[Andrew Burgess <aburgess@redhat.com>]

Simon Marchi <simark@simark.ca> writes:

> On 3/25/24 13:20, Andrew Burgess wrote:
>> Andrew Burgess <aburgess@redhat.com> writes:
>> 
>>> In v3:
>>>
>>>   - Rebased.  Nasty merge conflict with 4bb20a6244b7091 which I think
>>>     I've resolved, but am unable to test.  Reposting so the author of
>>>     that other commit can validate.
>>>
>>>   - Initial testing looks good.  Full tests are still running.
>> 
>> Testing completed with no issues.  H.J. Lu confirmed that this versions
>> didn't break the x32 behaviour.  I've gone ahead and pushed these
>> patches.
>> 
>> If anything crops up then do let me know.
>> 
>> Thanks,
>> Andrew
>> 
>
> Hi Andrew,
>
> I think your series introduces some build failures with Clang.  One is
> easy, it's a missing `-x c++` in gdbserver/Makefile.in.

Thanks for fixing this one.

>
> The other is:
>
>       CXX    nat/x86-linux-tdesc-ipa.o
>     /home/smarchi/src/binutils-gdb/gdbserver/../gdb/nat/x86-linux-tdesc.c:167:1: error: unused function 'x86_linux_i386_tdesc_feature_mask' [-Werror,-Wunused-function]
>       167 | x86_linux_i386_tdesc_feature_mask ()
>           | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
> It's possible that this function just needs to be moved in the same
> "#ifdef" as where it's used, but since I didn't follow your work
> closely, I prefer to let you fix it, in case I'm missing something.

Sorry for the breakage.

I pushed the patch below to resolve this issue.

Thanks,
Andrew

--

commit f4c19f89ef43dbce8065532c808e1aeb05d08994
Author: Andrew Burgess <aburgess@redhat.com>
Date:   Tue Mar 26 12:09:27 2024 +0000

    gdb/gdbserver: fix some defined but unused function warnings
    
    This commit:
    
      commit 198ff6ff819c240545f9fc68b39636fd376d4ba9
      Date:   Tue Jan 30 15:37:23 2024 +0000
    
          gdb/gdbserver: share x86/linux tdesc caching
    
    added some functions which are always defined, but their use is
    guarded within various #ifdef blocks.  As a result we were seeing
    errors about defined, but unused, functions.
    
    I've fixed this problem in this commit by wrapping the function
    definitions within #ifdef blocks.
    
    I'm a little worried that there might be too many #ifdef blocks within
    this file, however, I'm going to commit this fix for now as this will
    fix the build, then I'll think about if there's a better way to split
    this file so we might avoid some of these #ifdef blocks.

diff --git a/gdb/nat/x86-linux-tdesc.c b/gdb/nat/x86-linux-tdesc.c
index c438dfae84f..8a02f77fa6a 100644
--- a/gdb/nat/x86-linux-tdesc.c
+++ b/gdb/nat/x86-linux-tdesc.c
@@ -160,6 +160,8 @@ static constexpr x86_tdesc_feature x86_linux_all_tdesc_features[] = {
   { X86_XSTATE_X87,	true,	false, 	false }
 };
 
+#if defined __i386__ || !defined IN_PROCESS_AGENT
+
 /* Return a compile time constant which is a mask of all the cpu features
    that are checked for when building an i386 target description.  */
 
@@ -175,6 +177,10 @@ x86_linux_i386_tdesc_feature_mask ()
   return mask;
 }
 
+#endif /* __i386__ || !IN_PROCESS_AGENT */
+
+#ifdef __x86_64__
+
 /* Return a compile time constant which is a mask of all the cpu features
    that are checked for when building an amd64 target description.  */
 
@@ -205,6 +211,8 @@ x86_linux_x32_tdesc_feature_mask ()
   return mask;
 }
 
+#endif /* __x86_64__ */
+
 /* Return a compile time constant which is a count of the number of cpu
    features that are checked for when building an i386 target description.  */
 
@@ -222,6 +230,8 @@ x86_linux_i386_tdesc_count ()
   return (1 << count);
 }
 
+#if defined __x86_64__ || defined IN_PROCESS_AGENT
+
 /* Return a compile time constant which is a count of the number of cpu
    features that are checked for when building an amd64 target description.  */
 
@@ -256,6 +266,8 @@ x86_linux_x32_tdesc_count ()
   return (1 << count);
 }
 
+#endif /* __x86_64__ || IN_PROCESS_AGENT */
+
 #ifdef IN_PROCESS_AGENT
 
 /* See linux-x86-tdesc.h.  */