From 005159e92420a102516ee6e29ef2178c818da5d0 Mon Sep 17 00:00:00 2001
From: Alexey Samsonov <samsonov@google.com>
Date: Tue, 23 Apr 2013 10:17:34 +0000
Subject: [PATCH] Use zlib to uncompress debug sections in DWARF parser.

This makes llvm-dwarfdump and llvm-symbolizer understand
debug info sections compressed by ld.gold linker.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@180088 91177308-0d34-0410-b5e6-96231b3b80d8
---
 cmake/config-ix.cmake                         |   6 ++-
 lib/DebugInfo/DWARFContext.cpp                |  38 ++++++++++++++++++
 lib/DebugInfo/DWARFContext.h                  |   3 ++
 test/DebugInfo/Inputs/dwarfdump-test-zlib.cc  |  24 +++++++++++
 .../Inputs/dwarfdump-test-zlib.elf-x86-64     | Bin 0 -> 9589 bytes
 test/DebugInfo/dwarfdump-zlib.test            |  12 ++++++
 test/Makefile                                 |   1 +
 test/lit.cfg                                  |   3 ++
 test/lit.site.cfg.in                          |   1 +
 9 files changed, 87 insertions(+), 1 deletion(-)
 create mode 100644 test/DebugInfo/Inputs/dwarfdump-test-zlib.cc
 create mode 100755 test/DebugInfo/Inputs/dwarfdump-test-zlib.elf-x86-64
 create mode 100644 test/DebugInfo/dwarfdump-zlib.test

diff --git a/cmake/config-ix.cmake b/cmake/config-ix.cmake
index c426f41f5c9..7c4285a299b 100755
--- a/cmake/config-ix.cmake
+++ b/cmake/config-ix.cmake
@@ -105,7 +105,11 @@ if( NOT PURE_WINDOWS )
   endif()
   check_library_exists(dl dlopen "" HAVE_LIBDL)
   check_library_exists(rt clock_gettime "" HAVE_LIBRT)
-  check_library_exists(z compress2 "" HAVE_LIBZ)
+  if (LLVM_ENABLE_ZLIB)
+    check_library_exists(z compress2 "" HAVE_LIBZ)
+  else()
+    set(HAVE_LIBZ 0)
+  endif()
 endif()
 
 # function checks
diff --git a/lib/DebugInfo/DWARFContext.cpp b/lib/DebugInfo/DWARFContext.cpp
index 74a8c36d7ab..1e13731361a 100644
--- a/lib/DebugInfo/DWARFContext.cpp
+++ b/lib/DebugInfo/DWARFContext.cpp
@@ -10,6 +10,8 @@
 #include "DWARFContext.h"
 #include "llvm/ADT/SmallString.h"
 #include "llvm/ADT/StringSwitch.h"
+#include "llvm/ADT/STLExtras.h"
+#include "llvm/Support/Compression.h"
 #include "llvm/Support/Dwarf.h"
 #include "llvm/Support/Format.h"
 #include "llvm/Support/Path.h"
@@ -483,6 +485,22 @@ DIInliningInfo DWARFContext::getInliningInfoForAddress(uint64_t Address,
   return InliningInfo;
 }
 
+static bool consumeCompressedDebugSectionHeader(StringRef &data,
+                                                uint64_t &OriginalSize) {
+  // Consume "ZLIB" prefix.
+  if (!data.startswith("ZLIB"))
+    return false;
+  data = data.substr(4);
+  // Consume uncompressed section size (big-endian 8 bytes).
+  DataExtractor extractor(data, false, 8);
+  uint32_t Offset = 0;
+  OriginalSize = extractor.getU64(&Offset);
+  if (Offset == 0)
+    return false;
+  data = data.substr(Offset);
+  return true;
+}
+
 DWARFContextInMemory::DWARFContextInMemory(object::ObjectFile *Obj) :
   IsLittleEndian(Obj->isLittleEndian()),
   AddressSize(Obj->getBytesInAddress()) {
@@ -497,6 +515,22 @@ DWARFContextInMemory::DWARFContextInMemory(object::ObjectFile *Obj) :
 
     name = name.substr(name.find_first_not_of("._")); // Skip . and _ prefixes.
 
+    // Check if debug info section is compressed with zlib.
+    if (name.startswith("zdebug_")) {
+      uint64_t OriginalSize;
+      if (!zlib::isAvailable() ||
+          !consumeCompressedDebugSectionHeader(data, OriginalSize))
+        continue;
+      OwningPtr<MemoryBuffer> UncompressedSection;
+      if (zlib::uncompress(data, UncompressedSection, OriginalSize) !=
+          zlib::StatusOK)
+        continue;
+      // Make data point to uncompressed section contents and save its contents.
+      name = name.substr(1);
+      data = UncompressedSection->getBuffer();
+      UncompressedSections.push_back(UncompressedSection.take());
+    }
+
     StringRef *Section = StringSwitch<StringRef*>(name)
         .Case("debug_info", &InfoSection)
         .Case("debug_abbrev", &AbbrevSection)
@@ -584,4 +618,8 @@ DWARFContextInMemory::DWARFContextInMemory(object::ObjectFile *Obj) :
   }
 }
 
+DWARFContextInMemory::~DWARFContextInMemory() {
+  DeleteContainerPointers(UncompressedSections);
+}
+
 void DWARFContextInMemory::anchor() { }
diff --git a/lib/DebugInfo/DWARFContext.h b/lib/DebugInfo/DWARFContext.h
index 37b272993f3..78c18e61680 100644
--- a/lib/DebugInfo/DWARFContext.h
+++ b/lib/DebugInfo/DWARFContext.h
@@ -161,8 +161,11 @@ class DWARFContextInMemory : public DWARFContext {
   StringRef RangeDWOSection;
   StringRef AddrSection;
 
+  SmallVector<MemoryBuffer*, 4> UncompressedSections;
+
 public:
   DWARFContextInMemory(object::ObjectFile *);
+  ~DWARFContextInMemory();
   virtual bool isLittleEndian() const { return IsLittleEndian; }
   virtual uint8_t getAddressSize() const { return AddressSize; }
   virtual const RelocAddrMap &infoRelocMap() const { return InfoRelocMap; }
diff --git a/test/DebugInfo/Inputs/dwarfdump-test-zlib.cc b/test/DebugInfo/Inputs/dwarfdump-test-zlib.cc
new file mode 100644
index 00000000000..260c3c435c1
--- /dev/null
+++ b/test/DebugInfo/Inputs/dwarfdump-test-zlib.cc
@@ -0,0 +1,24 @@
+class DummyClass {
+  int a_;
+ public:
+  DummyClass(int a) : a_(a) {}
+  int add(int b) {
+    return a_ + b;
+  }
+};
+
+int f(int a, int b) {
+  DummyClass c(a);
+  return c.add(b);
+}
+
+int main() {
+  return f(2, 3);
+}
+
+// Built with Clang 3.2 and ld.gold linker:
+// $ mkdir -p /tmp/dbginfo
+// $ cp dwarfdump-test-zlib.cc /tmp/dbginfo
+// $ cd /tmp/dbginfo
+// $ clang++ -g dwarfdump-test-zlib.cc -Wl,--compress-debug-sections=zlib \
+//   -o <output>
diff --git a/test/DebugInfo/Inputs/dwarfdump-test-zlib.elf-x86-64 b/test/DebugInfo/Inputs/dwarfdump-test-zlib.elf-x86-64
new file mode 100755
index 0000000000000000000000000000000000000000..16b3153939fa1ba9e4765352951db65620d1c9a1
GIT binary patch
literal 9589
zcmeHNeQ;A%7QabbTA;L13X~5aK$I_=wtSQ#iYfgdEfGp7;%f1cHlYoqNljj`AS}g#
z8Viboh`1m)>$)PY0|<+YA4QOY<3~}{D&SOBBz%ns$aDeCo^#(l;qh7qcXsxVo#D>B
zckcQ4opbJaH}Bp1UKZrq^7J|#qo!wL7;?sF3l}M8#-?}WtdM20_RPWtvH`3eNTT7e
za1BzWN3=z_CK@N~qJSbUQ?Cdz^%mYC+Dj2LqLd-h7@<>$MQ{pD69I@agQ7+p=GxLp
z;ahI8@D5R<$RkZq!Xb+|V*)p4#@-g*AzCjIBTD1LvEflH{1xjhz#(rz!gBce3PoFN
zg-2xv_J4@wEKm#Y5Va{{MwD=q0Ec=0r>#ifu2q*fk$;4W0I6<QX~ytWw<E>v@&u|=
zsxvcEGKQP|Uh_~sZY(>NX554+96@zV1`cyhf&=FO>ZepE4*K$S)069~F4a7~ytDnl
zAJ->>s?CNzZ*D^~2xP`K<eAryGf9FGlu3Ts?vtfTyUWAI+49F`OGC{=%_A7<a4;EQ
zS6=D$2s?>M1+%3hXSvHSJAK)1yWj8hqp@Fhlnoi8Ab_om^vcU126E|2a+p7kA8G5R
z6k7+{KA;p+mj!yb+B2qSomGJ5J{Hd5H5Zpi(Ha#wP9Jp)YwfAH&;uY0oxsy#z*uW{
z$L$J_IMI`_@U5rdybN0SR^%(dh8=GXhr{Wo^Mm^zv<8n@YkxRXSe)NbU)9dStPMMx
zV~{jhfP@;r@cgo67{gi@69fm=#r3j2+{i=BfM!nYvj}~%Dc0c8$!J(}*DA1Mf$=*G
z@RI`Avj|&wG7h<&m!hn-Cw10QO@7mi#6CHosjb&pgE5t06i)h(8`Z@~2pQGnJF+Kc
zJaly!?ty8$)bXsfccpa~apjx0V+xpC+_d+TAyBxPTWj)z^bFRexo3j&F9!oHb%D#)
zU_&SykxS>FsSC7(%CTR&+c0g{G8_-uzmFH;J)F@ExtHP0LYtu|ijg`WvbP53w*&*{
zt#$Ju!1>S!Ktzny_SWD$Q%g`bfhBb6Mn*w5cxD$AMX9GokJ^z?fJnWqLAR;N7Ic`J
ztxHo9FTNkPExk1k3CKe>aD2vw5v;)y6I`uRnYKYkvx2|o2QOH6{u-4Z{5rJjR~W|P
z({iu$gNZQbY{84R;DwyvPg&ulL)O||aE(p|PFU+oOmOMlCWEccVM?|^&uFWYO#^L^
zbs7vTb5~-Y8Z65>ygI7<BJ%-eKt(q!LQ7cgz5{Q|;<}%VgLza3^Pwdi_!(UBE6nFw
z=uONGXEDxo<yV0B6>Pe>MNC(1(DjUujavpkgkB)S!}wb`oQein35Lh?*`4T~iMl)C
zm=5~BZRpW&2^>|RhcqJ%aWWh&kctVV8o!8J0mprS>w<dG&~_sn7%zyjNlRiLYbPaG
zYGa?)KNYn!T7O>xK1&dP0FJ$2r;HF7Xox=rhvjlOYy`>@40#F31&JLWj;mpJ_ZU5R
zn5iF%vFUn7jSc}G{q+L-128jhhkT=fmH;gQx&-qF`VP<>pof6Y19}<g3e*Gs9iZId
zjnj<^+^E2f3f!o`|7``}R?Zs@K+BUHS2Wl{V!cVvAY8{|qi11bl!f#7yuya-Pi#cD
z3q9?R2oDQe-iTK}BT+%m=T&+OvaBONGe-8b7i`g6xTb^9`vv{wx3CwxRU*Dl$Z7vX
z`~yNxyi>#DdK_E4LRd)n6Wd9)y`6<?&V-eCON1UCn>_vs!9$%nPjQ61|2d)ksy0y9
zXp2^)@Z++x?=TLWQX25cfmEBzWA}ND!_67yVJYbWE*vpP;g$UiUdL!Sa6eb>bvumd
z=5(>;Y12QBU(Y%*n<oBE+|SRg))~0Y(!tJ;_Uq(yFqd~`xFb-vOfFAixXM?zTwpM6
zl_oA;e;w=2l8qK_PS<CqkQ;@Z=5IB~(ND9+&mxdVv)1`vYg{jK|FEOg<oQ=FH?U^0
z{|A=0ZE)7Xs!%(w4?vFj)QEYgSTojy_WERI5c!k*D3^C;xcO4Ivs~UK@}8q!6yg#Z
z>06I0fnUhh@;)Hf>AOYl8%D%K8v*j0vHl`Yy&`70LT=R5(<bD&-NeS%%0Nqm9QU8t
zO#IB49ppy1JH%<m>jwEvERpRMaifL(BixT*RAed04Q=>Y4RWK(j=HaLKOH0Y1(LtX
z<q47d1If2u!_Pi$--#K-xcFWLK#d^p2X$K~<aFQtQOK)=ysrpwQON6sobIa$P;djx
zPnzeHPnT=R`wIC$O+8te-R)Mr8ewHV*)IoRXIBQVOtOn77D=}J$;A@9F_Chv;tH}=
zDu-7pcY90iZpk5geSXOvsAgr}$||>0b~?-%BZj3jR7fsIHFCq)44=KyDLDd_m2;Tm
zVY_bzbgNQir(aH)3ol;GWs&f?6LNU;yrQgoawXz0I`yvvFN*}1B!}NCRoFcax06yA
z_Bnr;kabUfHb`9_mn_+RKKmTyz0lPnr^mshakhzLvux7Dyu8V|#ZqzBSX(ZZ0^ci1
zW&VIP1AMRuIwD5TjK7U?*kwEE^Cv>j;qpiUzY~17dUFLytE^R51q;Akig<0jyprxu
zpW%X3?w^pJmV;&GZ?%Tm9gf^q((Lp$WJ6Kr!~`I-Qoo;)#xnf(?J~H4sSNivrDf_Z
zyhHDednw|`8Ew_=Q4qv}fbhk73s=*7StIAJoanwlk}Vp%thaELEF7B1P*5#?Cg5?+
zLh)DWE!=XI5KQNs@=hCbr{_zuknaGBYcRsM=q<p(`vxJPbq*K834N$rH3%-inuGAP
z4yqChJE|Y$P4US`9^i4!LU>x&ED?CJCp_5^Jq?a<4MaSxgH{N<Q3Q$sin)^?A!c`i
z5M_j?b>wz|r#cai@*(+X;d88TNbA(bHt@K%)8cWxfO)kOcv|;`G<czImA?fPYCNOo
zn@HG5Ktbu9FVWr_JUt(c0#EaZ+z^g<9|Tv&r*-OfF(C0-gUnmykA+7(y>qAM=U#!Q
zx|0LqiQ;n+@pSzf+gZ42qrfMjQq`#bB>EOQYCL@hq3<KKjwgGHOZ-%U7ndutex&am
z+Inf@7i#dd&Wsa<yB<EI!PB}eS>UztwESThJeG>`r*%AikE$2-!`}}!!c$%*kf90T
z>AOL5B3fUuSGd;vP*AAxv=0mkype!{QhXBQcOx~P)@5-bLCP~?PaQ$W14W;dKdoDm
ze<ywrU@?9g9O_2+A!;G?(H0xg{?MVmqEXJ%jS9uDDe9)RMP4gI`+~j->UD)g&RW|A
zpwY_E_!ohv{Wdw!+D3ywD?{7%{S?IY@XM}&zeV8ZQ(!^0vG!a8zxsE^e-Qj?ePXD;
z{Z9NYf!ALD4+Q=hbf#()i|AU=QD0@oCihpk>(y^EfN}kyZ8>88(({+}+WQxNH`kL=
zMW+jVllWoW20HviPrpBU{+$6l{dQDvmMIdyP2*fUR@`5-g)bN5$Nhx5Y4P{NM^#0y
zTVJhPTzFTaF3Ygyn}fduwx8=3y4?QNNhc0<>Ct}cz@!e72Ib^z-fCUZW9z-&r#|ui
zKR-FVwq*Ei-;CUH?z@BgPQ4IJ$}cXw$#tS&a894CJ@k2*B?tA{opr&vX8)<vRi9X$
zUw`*r)dR;bJbkFwq6=P&vs=z9(a-Hay6LT`?uJnfr$(>qcj83Sx))}3cI@9&bj#U!
zt}Q=a?DBTs`7_V<T2-)Q`ofo+m+txS-Fcr4Fx_*iPw%WvqxQYpG-X@fiK(;m-x#oW
z@fQaVbxVA2%$iG|f7Up3>b!+5y(fe}n!oSImr|#_Fs5JE+m~G#DGOKYraZMN;e)jJ
zoX<K9D*WQDzLR%%nDy9(D1FY3^_OGAJNLeNwrA7W#?AlwZq4cb6Jxt=d}!^rzGF50
z_WR{MYg#@Y6~F(Bm3um*b(QXWq9pCLsHWeZe*TeXrp{=$^7t`Z`MsTd?5mjW4I3xc
zY*;xxVenH^2GlIyJ~;KECpu)@yX|7!fjgLaQp1+jtD{!u{Qb=FM-tzBeA(>5hsV8f
z`q)j(TjlCzP^HSMb^P+`t|NQCTKD3|_4W0;*L5^?@4n#N&-(nlJVWfFqK`V)eO^@H
z?<kKisv9rA?i;%N{Ikm!ZjO(4$N$Nu-<U8xW7vTHJ-a=AWb(FSmJKfidseiw{(}up
z>5%kg*^$~O%{5Oxx#^*4<wq8E=xrJJ?#ey8<C0=n`Ml0q*GZ_Y#Ftm=a`!>qjY)am
zhbJWmDmEQ$zhraw*uy1zQ<C0k-}slprDNVbTr%P8(I)GtC%)dEw03i0W$Ntb#y&H0
z{{Ds?85^EoJM6=|4!zj*r>};O4BWo;UwOUnpVfEP>Q{Q|%kSQK=N+HLCw|%`m{}YD
z>8yg6CjYr;+k(>x3x@b64BPf-TK&so-foWV{Z;DOwi0LNxejM(pj@(-minBtnfYo?
zc2>*GT<(Pv?0zeq9)&YkxuN58y6sAjZ{^Ie#{xX=SixE$!g*DoRGw1>|40z-ukaMz
zXZMs}(c6g?YGRe-BHSF0U-m^t!25~??>Z|aWm8_^0Hc{@a4Po%%(I<7zYG320j$dd
zf0n3<IH<IjxjZwxlw^bzNvL&7pBKO}$8(iqX7+exCl7`18&gC)@Cxb9sbprp!{+)5
zyT3x1XN}EIk?rM?T;R_uPD(`S({NP&O9L(l;Bs6+dCMXo-ksyISGqvvuYi)s_EP8(
Pom0tqpO^2D&HVgt`TAbF

literal 0
HcmV?d00001

diff --git a/test/DebugInfo/dwarfdump-zlib.test b/test/DebugInfo/dwarfdump-zlib.test
new file mode 100644
index 00000000000..8ce2cf7690d
--- /dev/null
+++ b/test/DebugInfo/dwarfdump-zlib.test
@@ -0,0 +1,12 @@
+REQUIRES: zlib
+
+RUN: llvm-dwarfdump %p/Inputs/dwarfdump-test-zlib.elf-x86-64  \
+RUN:   | FileCheck %s -check-prefix FULLDUMP
+RUN: llvm-dwarfdump %p/Inputs/dwarfdump-test-zlib.elf-x86-64  \
+RUN:   --address=0x400559 --functions | FileCheck %s -check-prefix MAIN
+
+FULLDUMP: .debug_abbrev contents
+FULLDUMP: .debug_info contents
+
+MAIN: main
+MAIN-NEXT: /tmp/dbginfo{{[/\\]}}dwarfdump-test-zlib.cc:16
diff --git a/test/Makefile b/test/Makefile
index 8d05c57f862..2213319df81 100644
--- a/test/Makefile
+++ b/test/Makefile
@@ -141,6 +141,7 @@ lit.site.cfg: FORCE
 	@$(ECHOPATH) s=@LLVM_BINDINGS@=$(BINDINGS_TO_BUILD)=g >> lit.tmp
 	@$(ECHOPATH) s=@HOST_OS@=$(HOST_OS)=g >> lit.tmp
 	@$(ECHOPATH) s=@HOST_ARCH@=$(HOST_ARCH)=g >> lit.tmp
+	@$(ECHOPATH) s=@HAVE_LIBZ@=$(HAVE_LIBZ)=g >> lit.tmp
 	@sed -f lit.tmp $(PROJ_SRC_DIR)/lit.site.cfg.in > $@
 	@-rm -f lit.tmp
 
diff --git a/test/lit.cfg b/test/lit.cfg
index ea91f45754d..b423c6e80b7 100644
--- a/test/lit.cfg
+++ b/test/lit.cfg
@@ -268,6 +268,9 @@ if (config.llvm_use_sanitizer == "Memory" or
 if not 'hexagon' in config.target_triple:
     config.available_features.add("object-emission")
 
+if config.have_zlib == "1":
+    config.available_features.add("zlib")
+
 # llc knows whether he is compiled with -DNDEBUG.
 import subprocess
 try:
diff --git a/test/lit.site.cfg.in b/test/lit.site.cfg.in
index 8024b24fcde..3a680b28126 100644
--- a/test/lit.site.cfg.in
+++ b/test/lit.site.cfg.in
@@ -19,6 +19,7 @@ config.host_os = "@HOST_OS@"
 config.host_arch = "@HOST_ARCH@"
 config.llvm_use_intel_jitevents = "@LLVM_USE_INTEL_JITEVENTS@"
 config.llvm_use_sanitizer = "@LLVM_USE_SANITIZER@"
+config.have_zlib = "@HAVE_LIBZ@"
 
 # Support substitution of the tools_dir with user parameters. This is
 # used when we can't determine the tool dir at configuration time.