[PATCH] D43863: [ELF] - Do not remove empty sections that use symbols in expressions.

[Rafael Avila de Espindola via llvm-commits]

LGTM.

If we even decide to drop these sections something like this would be
needed to split ReferencedSymbols into a per CMD list.

Cheers,
Rafael

George Rimar via Phabricator via llvm-commits
<llvm-commits at lists.llvm.org> writes:

> grimar updated this revision to Diff 136265.
> grimar added a comment.
>
> - Updated testcase in according to review comments.
>
>
> https://reviews.llvm.org/D43863
>
> Files:
>   ELF/OutputSections.cpp
>   ELF/OutputSections.h
>   ELF/ScriptParser.cpp
>   test/ELF/linkerscript/empty-sections-expressions.s
>
>
> Index: test/ELF/linkerscript/empty-sections-expressions.s
> ===================================================================
> --- test/ELF/linkerscript/empty-sections-expressions.s
> +++ test/ELF/linkerscript/empty-sections-expressions.s
> @@ -0,0 +1,18 @@
> +# REQUIRES: x86
> +# RUN: llvm-mc -filetype=obj -triple=x86_64-unknown-linux %s -o %t.o
> +
> +## We remove empty sections that do not reference symbols in address,
> +## LMA, align and subalign expressions. Here we check that.
> +
> +# RUN: echo "SECTIONS { .debug_info 0 : { *(.debug_info) } }" > %t.script
> +# RUN: ld.lld -o %t --script %t.script %t.o
> +# RUN: llvm-objdump -section-headers %t | FileCheck %s
> +# CHECK-NOT: .debug_info
> +
> +# RUN: echo "SECTIONS { .debug_info foo : { *(.debug_info) } }" > %t2.script
> +# RUN: ld.lld -o %t2 --script %t2.script %t.o
> +# RUN: llvm-objdump -section-headers %t2 | FileCheck %s --check-prefix=SEC
> +# SEC: .debug_info
> +
> +.globl foo
> +foo:
> Index: ELF/ScriptParser.cpp
> ===================================================================
> --- ELF/ScriptParser.cpp
> +++ ELF/ScriptParser.cpp
> @@ -668,6 +668,8 @@
>    OutputSection *Cmd =
>        Script->createOutputSection(OutSec, getCurrentLocation());
>  
> +  size_t SymbolsReferenced = Script->ReferencedSymbols.size();
> +
>    if (peek() != ":")
>      readSectionAddressType(Cmd);
>    expect(":");
> @@ -734,6 +736,8 @@
>    // Consume optional comma following output section command.
>    consume(",");
>  
> +  if (Script->ReferencedSymbols.size() > SymbolsReferenced)
> +    Cmd->ExpressionsUseSymbols = true;
>    return Cmd;
>  }
>  
> Index: ELF/OutputSections.h
> ===================================================================
> --- ELF/OutputSections.h
> +++ ELF/OutputSections.h
> @@ -103,6 +103,7 @@
>    std::string LMARegionName;
>    bool NonAlloc = false;
>    bool Noload = false;
> +  bool ExpressionsUseSymbols = false;
>  
>    template <class ELFT> void finalize();
>    template <class ELFT> void writeTo(uint8_t *Buf);
> Index: ELF/OutputSections.cpp
> ===================================================================
> --- ELF/OutputSections.cpp
> +++ ELF/OutputSections.cpp
> @@ -84,11 +84,10 @@
>    if (!Phdrs.empty())
>      return false;
>  
> -  // We do not want to remove sections that have custom address or align
> -  // expressions set even if them are empty. We keep them because we
> -  // want to be sure that any expressions can be evaluated and report
> -  // an error otherwise.
> -  if (AddrExpr || AlignExpr || LMAExpr)
> +  // We do not want to remove sections that reference symbols in address and
> +  // other expressions. We add script symbols as undefined, and want to ensure
> +  // all of them are defined in output, hence have to keep them.
> +  if (ExpressionsUseSymbols)
>      return false;
>  
>    for (BaseCommand *Base : SectionCommands)
>
>
> Index: test/ELF/linkerscript/empty-sections-expressions.s
> ===================================================================
> --- test/ELF/linkerscript/empty-sections-expressions.s
> +++ test/ELF/linkerscript/empty-sections-expressions.s
> @@ -0,0 +1,18 @@
> +# REQUIRES: x86
> +# RUN: llvm-mc -filetype=obj -triple=x86_64-unknown-linux %s -o %t.o
> +
> +## We remove empty sections that do not reference symbols in address,
> +## LMA, align and subalign expressions. Here we check that.
> +
> +# RUN: echo "SECTIONS { .debug_info 0 : { *(.debug_info) } }" > %t.script
> +# RUN: ld.lld -o %t --script %t.script %t.o
> +# RUN: llvm-objdump -section-headers %t | FileCheck %s
> +# CHECK-NOT: .debug_info
> +
> +# RUN: echo "SECTIONS { .debug_info foo : { *(.debug_info) } }" > %t2.script
> +# RUN: ld.lld -o %t2 --script %t2.script %t.o
> +# RUN: llvm-objdump -section-headers %t2 | FileCheck %s --check-prefix=SEC
> +# SEC: .debug_info
> +
> +.globl foo
> +foo:
> Index: ELF/ScriptParser.cpp
> ===================================================================
> --- ELF/ScriptParser.cpp
> +++ ELF/ScriptParser.cpp
> @@ -668,6 +668,8 @@
>    OutputSection *Cmd =
>        Script->createOutputSection(OutSec, getCurrentLocation());
>  
> +  size_t SymbolsReferenced = Script->ReferencedSymbols.size();
> +
>    if (peek() != ":")
>      readSectionAddressType(Cmd);
>    expect(":");
> @@ -734,6 +736,8 @@
>    // Consume optional comma following output section command.
>    consume(",");
>  
> +  if (Script->ReferencedSymbols.size() > SymbolsReferenced)
> +    Cmd->ExpressionsUseSymbols = true;
>    return Cmd;
>  }
>  
> Index: ELF/OutputSections.h
> ===================================================================
> --- ELF/OutputSections.h
> +++ ELF/OutputSections.h
> @@ -103,6 +103,7 @@
>    std::string LMARegionName;
>    bool NonAlloc = false;
>    bool Noload = false;
> +  bool ExpressionsUseSymbols = false;
>  
>    template <class ELFT> void finalize();
>    template <class ELFT> void writeTo(uint8_t *Buf);
> Index: ELF/OutputSections.cpp
> ===================================================================
> --- ELF/OutputSections.cpp
> +++ ELF/OutputSections.cpp
> @@ -84,11 +84,10 @@
>    if (!Phdrs.empty())
>      return false;
>  
> -  // We do not want to remove sections that have custom address or align
> -  // expressions set even if them are empty. We keep them because we
> -  // want to be sure that any expressions can be evaluated and report
> -  // an error otherwise.
> -  if (AddrExpr || AlignExpr || LMAExpr)
> +  // We do not want to remove sections that reference symbols in address and
> +  // other expressions. We add script symbols as undefined, and want to ensure
> +  // all of them are defined in output, hence have to keep them.
> +  if (ExpressionsUseSymbols)
>      return false;
>  
>    for (BaseCommand *Base : SectionCommands)
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits