public inbox for binutils@sourceware.org
 help / color / mirror / Atom feed
From: liuzhensong <liuzhensong@loongson.cn>
To: WANG Xuerui <i.swmail@xen0n.name>, binutils@sourceware.org
Cc: xry111@xry111.site, maskray@google.com, mengqinggang@loongson.cn
Subject: Re: [PATCH v2] LoongArch: Set e_flags to 0x40 of ELF while using new relocations.
Date: Tue, 27 Sep 2022 10:11:35 +0800	[thread overview]
Message-ID: <1a59e0da-fab2-9aab-c8de-e0a27d82341d@loongson.cn> (raw)
In-Reply-To: <83e9a5ac-c509-066b-07f6-cb3ec4583e48@xen0n.name>


在 2022/9/25 下午4:28, WANG Xuerui 写道:
> Hi,
>
> Overall this is going in the right direction (finally!) and 
> implementing the correct behavior, but a few small nits remain.
>
> Also, @xry111 WDYT?
>
>
> On 2022/9/21 09:17, liuzhensong wrote:
>
>>    As the docs[1] describes, update the e_flags of ELF.
>>    1. 32 or 64 bit is represented by EI_CLASS.
>>    2. The bits of e_flags
>>         Base ABI [2:0]:
>>           soft float    --  0x1
>>           single float  --  0x2
>>           double float  --  0x3
>>
>>         ABI version [7:6]:
>>           obj-v0 -- 0x0
>>           obj-v1 -- 0x1
>>    3. Discussions [2].
>>
>>    docs:
>>    [1] https://github.com/loongson/LoongArch-Documentation/blob/
>> main/docs/LoongArch-ELF-ABI-EN.adoc#e_flags-identifies-abi-type-and-version
>>    [2] https://github.com/loongson/LoongArch-Documentation/pull/61
>
> Rewritten into (somewhat more) proper English for you, also we have 
> another significant discussion thread that you hadn't mentioned:
>
> ```
> Subject: LoongArch: Update ELF e_flags handling according to 
> specification.
>
> Update handling of e_flags according to the documentation update [1]
> (discussions [2][3]).
>
> Object file bitness is now represented in the EI_CLASS byte. The e_flags
> field is now interpreted as follows:
>
> e_flags[2:0]: Base ABI modifier
>
> - 0x1: soft-float
> - 0x2: single-precision hard-float
> - 0x3: double-precision hard-float
>
> e_flags[7:6]: ELF object ABI version
>
> - 0x0: v0
> - 0x1: v1
>
> [1]: 
> https://github.com/loongson/LoongArch-Documentation/blob/main/docs/LoongArch-ELF-ABI-EN.adoc#e_flags-identifies-abi-type-and-version
> [2]: https://github.com/loongson/LoongArch-Documentation/pull/61
> [3]: https://github.com/loongson/LoongArch-Documentation/pull/47
> ```
>
> (Tip: next time when writing commit messages (or in fact anything) in 
> English, please bear in mind that Chinese is a "topic-prominent" 
> language [99] unlike English, meaning the idiomatic sentence pattern 
> is usually different. In general people can still grasp the meaning 
> but will probably have a headache in trying to do so...)
>
> [99]: https://en.wikipedia.org/wiki/Topic-prominent_language
>
>> ---
>>   bfd/elfnn-loongarch.c     | 21 ++++++++++++++++---
>>   binutils/readelf.c        | 10 ++++-----
>>   gas/config/tc-loongarch.c | 20 +++++++++---------
>>   include/elf/loongarch.h   | 44 +++++++++++++++++++--------------------
>>   4 files changed, 54 insertions(+), 41 deletions(-)
>>
>> diff --git a/bfd/elfnn-loongarch.c b/bfd/elfnn-loongarch.c
>> index 33c85e5207c..5b4b83eb07a 100644
>> --- a/bfd/elfnn-loongarch.c
>> +++ b/bfd/elfnn-loongarch.c
>> @@ -416,8 +416,22 @@ elfNN_loongarch_merge_private_bfd_data (bfd 
>> *ibfd, struct bfd_link_info *info)
>>         elf_elfheader (obfd)->e_flags = in_flags;
>>         return true;
>>       }
>> +  else if (out_flags != in_flags)
>> +    {
>> +      if ((EF_LOONGARCH_IS_OBJ_V0 (out_flags)
>> +       && EF_LOONGARCH_IS_OBJ_V1 (in_flags))
>> +      || (EF_LOONGARCH_IS_OBJ_V0 (in_flags)
>> +          && EF_LOONGARCH_IS_OBJ_V1 (out_flags)))
>> +    {
>> +      elf_elfheader (obfd)->e_flags |= EF_LOONGARCH_ABI_OBJ_V1;
>> +      out_flags = elf_elfheader (obfd)->e_flags;
>> +      in_flags = out_flags;
>> +    }
>> +    }
>>       /* Disallow linking different ABIs.  */
>> +  /* Only check relocation version.
>> +     The obj_v0 is compatible with obj_v1 in this "release cycle".  */
> We know this cycle is 2.40 so directly mention that?
>>     if (EF_LOONGARCH_ABI(out_flags ^ in_flags) & EF_LOONGARCH_ABI_MASK)
>>       {
>>         _bfd_error_handler (_("%pB: can't link different ABI 
>> object."), ibfd);
>> @@ -1579,15 +1593,16 @@ loongarch_elf_size_dynamic_sections (bfd 
>> *output_bfd,
>>         if (bfd_link_executable (info) && !info->nointerp)
>>       {
>>         const char *interpreter;
>> -      flagword flags = elf_elfheader (output_bfd)->e_flags;
>>         s = bfd_get_linker_section (dynobj, ".interp");
>>         BFD_ASSERT (s != NULL);
>> -      if (EF_LOONGARCH_IS_ILP32 (flags))
>> +
>> +      if (elf_elfheader (output_bfd)->e_ident[EI_CLASS] == ELFCLASS32)
>>           interpreter = "/lib32/ld.so.1";
>> -      else if (EF_LOONGARCH_IS_LP64 (flags))
>> +      else if (elf_elfheader (output_bfd)->e_ident[EI_CLASS] == 
>> ELFCLASS64)
>>           interpreter = "/lib64/ld.so.1";
>>         else
>>           interpreter = "/lib/ld.so.1";
>> +
> Wait what, are we still carrying the old-world defaults for the 
> interpreter path here? I know it's harmless because in fact it gets 
> overridden by gcc and glibc, but I'd suggest fixing it in a separate 
> patch later for consistency.
>>         s->contents = (unsigned char *) interpreter;
>>         s->size = strlen (interpreter) + 1;
>>       }
>> diff --git a/binutils/readelf.c b/binutils/readelf.c
>> index cafba9a4f56..6926cf16140 100644
>> --- a/binutils/readelf.c
>> +++ b/binutils/readelf.c
>> @@ -4340,11 +4340,6 @@ get_machine_flags (Filedata * filedata, 
>> unsigned e_flags, unsigned e_machine)
>>           }
>>         break;
>>       case EM_LOONGARCH:
>> -      if (EF_LOONGARCH_IS_LP64 (e_flags))
>> -        strcat (buf, ", LP64");
>> -      else if (EF_LOONGARCH_IS_ILP32 (e_flags))
>> -        strcat (buf, ", ILP32");
>> -
>>         if (EF_LOONGARCH_IS_SOFT_FLOAT (e_flags))
>>           strcat (buf, ", SOFT-FLOAT");
>>         else if (EF_LOONGARCH_IS_SINGLE_FLOAT (e_flags))
>> @@ -4352,6 +4347,11 @@ get_machine_flags (Filedata * filedata, 
>> unsigned e_flags, unsigned e_machine)
>>         else if (EF_LOONGARCH_IS_DOUBLE_FLOAT (e_flags))
>>           strcat (buf, ", DOUBLE-FLOAT");
>>   +      if (EF_LOONGARCH_IS_OBJ_V0 (e_flags))
>> +        strcat (buf, ", OBJ-v0");
>> +      else if (EF_LOONGARCH_IS_OBJ_V1 (e_flags))
>> +        strcat (buf, ", OBJ-v1");
>> +
>>         break;
>>       }
>>       }
>> diff --git a/gas/config/tc-loongarch.c b/gas/config/tc-loongarch.c
>> index 5bb8d317c5a..3e9204a5fa4 100644
>> --- a/gas/config/tc-loongarch.c
>> +++ b/gas/config/tc-loongarch.c
>> @@ -143,13 +143,13 @@ md_parse_option (int c, const char *arg)
>>     char ilp32[256] = "";
>>     unsigned char *suf = (unsigned char *)arg;
>>   -  lp64['s'] = lp64['S'] = EF_LOONGARCH_ABI_LP64_SOFT_FLOAT;
>> -  lp64['f'] = lp64['F'] = EF_LOONGARCH_ABI_LP64_SINGLE_FLOAT;
>> -  lp64['d'] = lp64['D'] = EF_LOONGARCH_ABI_LP64_DOUBLE_FLOAT;
>> +  lp64['s'] = lp64['S'] = EF_LOONGARCH_ABI_SOFT_FLOAT;
>> +  lp64['f'] = lp64['F'] = EF_LOONGARCH_ABI_SINGLE_FLOAT;
>> +  lp64['d'] = lp64['D'] = EF_LOONGARCH_ABI_DOUBLE_FLOAT;
>>   -  ilp32['s'] = ilp32['S'] = EF_LOONGARCH_ABI_ILP32_SOFT_FLOAT;
>> -  ilp32['f'] = ilp32['F'] = EF_LOONGARCH_ABI_ILP32_SINGLE_FLOAT;
>> -  ilp32['d'] = ilp32['D'] = EF_LOONGARCH_ABI_ILP32_DOUBLE_FLOAT;
>> +  ilp32['s'] = ilp32['S'] = EF_LOONGARCH_ABI_SOFT_FLOAT;
>> +  ilp32['f'] = ilp32['F'] = EF_LOONGARCH_ABI_SINGLE_FLOAT;
>> +  ilp32['d'] = ilp32['D'] = EF_LOONGARCH_ABI_DOUBLE_FLOAT;
>>       switch (c)
>>       {
>> @@ -216,24 +216,24 @@ void
>>   loongarch_after_parse_args ()
>>   {
>>     /* Set default ABI/ISA LP64D.  */
>> -  if (!EF_LOONGARCH_IS_LP64(LARCH_opts.ase_abi)
>> -      && !EF_LOONGARCH_IS_ILP32(LARCH_opts.ase_abi))
>> +  if (!LARCH_opts.ase_ilp32)
>>       {
>>         if (strcmp (default_arch, "loongarch64") == 0)
>>       {
>> -      LARCH_opts.ase_abi = EF_LOONGARCH_ABI_LP64_DOUBLE_FLOAT;
>> +      LARCH_opts.ase_abi = EF_LOONGARCH_ABI_DOUBLE_FLOAT;
>>         LARCH_opts.ase_ilp32 = 1;
>>         LARCH_opts.ase_lp64 = 1;
>>       }
>>         else if (strcmp (default_arch, "loongarch32") == 0)
>>       {
>> -      LARCH_opts.ase_abi = EF_LOONGARCH_ABI_ILP32_DOUBLE_FLOAT;
>> +      LARCH_opts.ase_abi = EF_LOONGARCH_ABI_DOUBLE_FLOAT;
>>         LARCH_opts.ase_ilp32 = 1;
>>       }
>>         else
>>       as_bad ("unknown default architecture `%s'", default_arch);
>>       }
>>   +  LARCH_opts.ase_abi |= EF_LOONGARCH_ABI_OBJ_V1;
>>     /* Set default ISA double-float.  */
>>     if (!LARCH_opts.ase_nf
>>         && !LARCH_opts.ase_sf
>> diff --git a/include/elf/loongarch.h b/include/elf/loongarch.h
>> index 74757b82ca8..cceaa8cccc4 100644
>> --- a/include/elf/loongarch.h
>> +++ b/include/elf/loongarch.h
>> @@ -232,36 +232,34 @@ RELOC_NUMBER (R_LARCH_RELAX, 100)
>>   END_RELOC_NUMBERS (R_LARCH_count)
>>     /* Processor specific flags for the ELF header e_flags field.  */
>> -/*The flag lp64s/lp64f/lp64d/ilp32s/ilp32f/ilp32d 3bits. */
>> -#define EF_LOONGARCH_ABI_LP64_SOFT_FLOAT    0x1
>> -#define EF_LOONGARCH_ABI_LP64_SINGLE_FLOAT    0x2
>> -#define EF_LOONGARCH_ABI_LP64_DOUBLE_FLOAT    0x3
>> +/* ase ABI modifier, 3bits.  */
> Oh please no more MIPS "ASE"s. It's "Base ABI modifier" in the docs...
I will fixed it in next patch.
>> +#define EF_LOONGARCH_ABI_SOFT_FLOAT 0x1
>> +#define EF_LOONGARCH_ABI_SINGLE_FLOAT    0x2
>> +#define EF_LOONGARCH_ABI_DOUBLE_FLOAT    0x3
>> +#define EF_LOONGARCH_ABI_MODIFIER_MASK    0x7
>>   -#define EF_LOONGARCH_ABI_ILP32_SOFT_FLOAT    0x5
>> -#define EF_LOONGARCH_ABI_ILP32_SINGLE_FLOAT    0x6
>> -#define EF_LOONGARCH_ABI_ILP32_DOUBLE_FLOAT    0x7
>> +#define EF_LOONGARCH_ABI_OBJ_V1      0x40
>> +#define EF_LOONGARCH_ABI_OBJ_MASK    0xC0
> I'd bikeshed a little and suggest "EF_LOONGARCH_OBJABI_foo" because 
> it's actually "ELF object file ABI v1" and not "ELF ABI object v1".
>>   -#define EF_LOONGARCH_ABI_MASK                0x7
>> -#define EF_LOONGARCH_ABI_ILP32_MASK            0x4
>> -#define EF_LOONGARCH_ABI_FLOAT_MASK            0x3
>> -#define EF_LOONGARCH_ABI_SOFT_FLOAT_MASK    0x1
>> -#define EF_LOONGARCH_ABI_SINGLE_FLOAT_MASK    0x2
>> -#define EF_LOONGARCH_ABI_DOUBLE_FLOAT_MASK    0x3
>> +#define EF_LOONGARCH_ABI_MASK \
>> +      (EF_LOONGARCH_ABI_OBJ_MASK | EF_LOONGARCH_ABI_MODIFIER_MASK)
>>   -#define EF_LOONGARCH_ABI(abi)    (EF_LOONGARCH_ABI_MASK & (abi))
>> +#define EF_LOONGARCH_ABI_MODIFIER(abi) \
>> +      (EF_LOONGARCH_ABI_MODIFIER_MASK & (abi))
>> +#define EF_LOONGARCH_ABI_OBJ(abi) \
>> +      (EF_LOONGARCH_ABI_OBJ_MASK & (abi))
>>   -#define EF_LOONGARCH_IS_LP64(abi) \
>> -  (EF_LOONGARCH_ABI(abi) && (!(EF_LOONGARCH_ABI(abi) & 
>> EF_LOONGARCH_ABI_ILP32_MASK)))
>> -#define EF_LOONGARCH_IS_ILP32(abi) \
>> -  (EF_LOONGARCH_ABI(abi) && (EF_LOONGARCH_ABI(abi) & 
>> EF_LOONGARCH_ABI_ILP32_MASK))
>> +#define EF_LOONGARCH_ABI(abi) ((abi) & EF_LOONGARCH_ABI_MASK)
>>     #define EF_LOONGARCH_IS_SOFT_FLOAT(abi) \
>> -  (!((EF_LOONGARCH_ABI(abi) & EF_LOONGARCH_ABI_FLOAT_MASK) ^ 
>> EF_LOONGARCH_ABI_SOFT_FLOAT_MASK))
>> -
>> +  (EF_LOONGARCH_ABI_MODIFIER (abi) == EF_LOONGARCH_ABI_SOFT_FLOAT)
>>   #define EF_LOONGARCH_IS_SINGLE_FLOAT(abi) \
>> -  (!((EF_LOONGARCH_ABI(abi) & EF_LOONGARCH_ABI_FLOAT_MASK) ^ 
>> EF_LOONGARCH_ABI_SINGLE_FLOAT_MASK))
>> -
>> +  (EF_LOONGARCH_ABI_MODIFIER (abi) == EF_LOONGARCH_ABI_SINGLE_FLOAT)
>>   #define EF_LOONGARCH_IS_DOUBLE_FLOAT(abi) \
>> -  (!((EF_LOONGARCH_ABI(abi) & EF_LOONGARCH_ABI_FLOAT_MASK) ^ 
>> EF_LOONGARCH_ABI_DOUBLE_FLOAT_MASK))
>> +  (EF_LOONGARCH_ABI_MODIFIER (abi) == EF_LOONGARCH_ABI_DOUBLE_FLOAT)
>> +
>> +#define EF_LOONGARCH_IS_OBJ_V0(abi) (!EF_LOONGARCH_ABI_OBJ(abi))
>> +#define EF_LOONGARCH_IS_OBJ_V1(abi) \
>> +      (EF_LOONGARCH_ABI_OBJ (abi) == EF_LOONGARCH_ABI_OBJ_V1)
>>     #endif /* _ELF_LOONGARCH_H */


      parent reply	other threads:[~2022-09-27  2:11 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-09-21  1:17 liuzhensong
2022-09-25  8:28 ` WANG Xuerui
2022-09-25  9:22   ` Xi Ruoyao
2022-09-27 16:49     ` Fangrui Song
2022-09-27  2:11   ` liuzhensong [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1a59e0da-fab2-9aab-c8de-e0a27d82341d@loongson.cn \
    --to=liuzhensong@loongson.cn \
    --cc=binutils@sourceware.org \
    --cc=i.swmail@xen0n.name \
    --cc=maskray@google.com \
    --cc=mengqinggang@loongson.cn \
    --cc=xry111@xry111.site \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).