Add some basic DWARF4 support. Based on patch from Jakub Jelinek

but with support for VLIW architectures with multiple opcodes per
instruction removed. Fixes #233595.


git-svn-id: svn://svn.valgrind.org/valgrind/trunk@11106 a5019735-40e9-0310-863c-91ae7b9d1cf9

4 files changed, 200 insertions, 24 deletions

diff --git a/coregrind/m_debuginfo/d3basics.c b/coregrind/m_debuginfo/d3basics.c
index e4d4ed1c..160ec2db 100644
--- a/coregrind/m_debuginfo/d3basics.c
+++ b/coregrind/m_debuginfo/d3basics.c
@@ -124,6 +124,10 @@ HChar* ML_(pp_DW_TAG) ( DW_TAG tag )
       case DW_TAG_imported_unit:      return "DW_TAG_imported_unit";
       case DW_TAG_condition:          return "DW_TAG_condition";
       case DW_TAG_shared_type:        return "DW_TAG_shared_type";
+      /* DWARF 4.  */
+      case DW_TAG_type_unit:          return "DW_TAG_type_unit";
+      case DW_TAG_rvalue_reference_type: return "DW_TAG_rvalue_reference_type";
+      case DW_TAG_template_alias:     return "DW_TAG_template_alias";
       /* SGI/MIPS Extensions.  */
       case DW_TAG_MIPS_loop:          return "DW_TAG_MIPS_loop";
       /* HP extensions.  See:
@@ -172,6 +176,10 @@ HChar* ML_(pp_DW_FORM) ( DW_FORM form )
       case DW_FORM_ref8:      return "DW_FORM_ref8";
       case DW_FORM_ref_udata: return "DW_FORM_ref_udata";
       case DW_FORM_indirect:  return "DW_FORM_indirect";
+      case DW_FORM_sec_offset:return "DW_FORM_sec_offset";
+      case DW_FORM_exprloc:   return "DW_FORM_exprloc";
+      case DW_FORM_flag_present:return "DW_FORM_flag_present";
+      case DW_FORM_ref_sig8:  return "DW_FORM_ref_sig8";
       default:                return "DW_FORM_???";
    }
 }
@@ -269,6 +277,13 @@ HChar* ML_(pp_DW_AT) ( DW_AT attr )
       case DW_AT_elemental: return "DW_AT_elemental";
       case DW_AT_pure: return "DW_AT_pure";
       case DW_AT_recursive: return "DW_AT_recursive";
+      /* DWARF 4 values.  */
+      case DW_AT_signature: return "DW_AT_signature";
+      case DW_AT_main_subprogram: return "DW_AT_main_subprogram";
+      case DW_AT_data_bit_offset: return "DW_AT_data_bit_offset";
+      case DW_AT_const_expr: return "DW_AT_const_expr";
+      case DW_AT_enum_class: return "DW_AT_enum_class";
+      case DW_AT_linkage_name: return "DW_AT_linkage_name";
       /* SGI/MIPS extensions.  */
       /* case DW_AT_MIPS_fde: return "DW_AT_MIPS_fde"; */
       /* DW_AT_MIPS_fde == DW_AT_HP_unmodifiable */
diff --git a/coregrind/m_debuginfo/priv_d3basics.h b/coregrind/m_debuginfo/priv_d3basics.h
index 81f46ef9..f14f178f 100644
--- a/coregrind/m_debuginfo/priv_d3basics.h
+++ b/coregrind/m_debuginfo/priv_d3basics.h
@@ -104,6 +104,10 @@ typedef enum
     DW_TAG_imported_unit = 0x3d,
     DW_TAG_condition = 0x3f,
     DW_TAG_shared_type = 0x40,
+    /* DWARF 4.  */
+    DW_TAG_type_unit = 0x41,
+    DW_TAG_rvalue_reference_type = 0x42,
+    DW_TAG_template_alias = 0x43,
     /* SGI/MIPS Extensions.  */
     DW_TAG_MIPS_loop = 0x4081,
     /* HP extensions.  See: ftp://ftp.hp.com/pub/lang/tools/WDB/wdb-4.0.tar.gz .  */
@@ -158,6 +162,8 @@ typedef enum dwarf_source_language
     DW_LANG_ObjC_plus_plus = 0x0011,
     DW_LANG_UPC = 0x0012,
     DW_LANG_D = 0x0013,
+    /* DWARF 4.  */
+    DW_LANG_Python = 0x0014,
     /* MIPS.  */
     DW_LANG_Mips_Assembler = 0x8001,
     /* UPC.  */
@@ -188,7 +194,12 @@ typedef enum
     DW_FORM_ref4 = 0x13,
     DW_FORM_ref8 = 0x14,
     DW_FORM_ref_udata = 0x15,
-    DW_FORM_indirect = 0x16
+    DW_FORM_indirect = 0x16,
+    /* DWARF 4 values.  */
+    DW_FORM_sec_offset = 0x17,
+    DW_FORM_exprloc = 0x18,
+    DW_FORM_flag_present = 0x19,
+    DW_FORM_ref_sig8 = 0x20
   }
   DW_FORM;
 
@@ -285,6 +296,13 @@ typedef enum
     DW_AT_elemental     = 0x66,
     DW_AT_pure          = 0x67,
     DW_AT_recursive     = 0x68,
+    /* DWARF 4 values.  */
+    DW_AT_signature       = 0x69,
+    DW_AT_main_subprogram = 0x6a,
+    DW_AT_data_bit_offset = 0x6b,
+    DW_AT_const_expr      = 0x6c,
+    DW_AT_enum_class      = 0x6d,
+    DW_AT_linkage_name    = 0x6e,
     /* SGI/MIPS extensions.  */
     DW_AT_MIPS_fde = 0x2001,
     DW_AT_MIPS_loop_begin = 0x2002,
diff --git a/coregrind/m_debuginfo/readdwarf.c b/coregrind/m_debuginfo/readdwarf.c
index d4204806..8c353854 100644
--- a/coregrind/m_debuginfo/readdwarf.c
+++ b/coregrind/m_debuginfo/readdwarf.c
@@ -1,6 +1,6 @@
 
 /*--------------------------------------------------------------------*/
-/*--- Read DWARF1/2/3 debug info.                      readdwarf.c ---*/
+/*--- Read DWARF1/2/3/4 debug info.                    readdwarf.c ---*/
 /*--------------------------------------------------------------------*/
 
 /*
@@ -139,6 +139,7 @@ typedef struct
   UShort li_version;
   ULong  li_header_length;
   UChar  li_min_insn_length;
+  UChar  li_max_ops_per_insn;
   UChar  li_default_is_stmt;
   Int    li_line_base;
   UChar  li_line_range;
@@ -182,7 +183,8 @@ enum dwarf_line_number_x_ops
   {
     DW_LNE_end_sequence = 1,
     DW_LNE_set_address = 2,
-    DW_LNE_define_file = 3
+    DW_LNE_define_file = 3,
+    DW_LNE_set_discriminator = 4
   };
 
 typedef struct
@@ -199,7 +201,7 @@ typedef struct
   UInt  column;
   Int   is_stmt;
   Int   basic_block;
-  Int   end_sequence;
+  UChar end_sequence;
 } LineSMR;
 
 
@@ -411,6 +413,11 @@ Word process_extended_line_op( struct _DebugInfo* di,
             VG_(printf)("  DWARF2-line: set_address\n");
          break;
 
+      case DW_LNE_set_discriminator:
+         read_leb128 (data, & bytes_read, 0);
+         data += bytes_read;
+         break;
+
       default:
          if (di->ddump_line)
             VG_(printf)("process_extended_line_op:default\n");
@@ -513,9 +520,9 @@ void read_dwarf2_lineblock ( struct _DebugInfo* di,
       VG_(printf)("  DWARF Version:               %d\n", 
                   (Int)info.li_version);
 
-   if (info.li_version != 2 && info.li_version != 3) {
+   if (info.li_version != 2 && info.li_version != 3 && info.li_version != 4) {
       ML_(symerr)(di, True,
-                  "Only DWARF version 2 and 3 line info "
+                  "Only DWARF version 2, 3 and 4 line info "
                   "is currently supported.");
       goto out;
    }
@@ -533,6 +540,26 @@ void read_dwarf2_lineblock ( struct _DebugInfo* di,
       VG_(printf)("  Minimum Instruction Length:  %d\n", 
                   (Int)info.li_min_insn_length);
 
+   /* We only support machines with one opcode per instruction
+      for now. If we ever want to support VLIW machines there is
+      code to handle multiple opcodes per instruction in the
+      patch attached to BZ#233595.
+   */
+   if (info.li_version >= 4) {
+      info.li_max_ops_per_insn = * ((UChar *)external);
+      if (info.li_max_ops_per_insn != 1) {
+         ML_(symerr)(di, True,
+                     "Invalid Maximum Ops Per Insn in line info.");
+         goto out;
+      }
+      external += 1;
+      if (di->ddump_line)
+         VG_(printf)("  Maximum Ops Per Insn:        %d\n", 
+                  (Int)info.li_max_ops_per_insn);
+   } else {
+      info.li_max_ops_per_insn = 1;
+   }
+
    info.li_default_is_stmt = * ((UChar *)external);
    external += 1;
    if (di->ddump_line)
@@ -714,7 +741,7 @@ void read_dwarf2_lineblock ( struct _DebugInfo* di,
 
          Int advAddr;
          op_code -= info.li_opcode_base;
-         adv      = (op_code / info.li_line_range) 
+         adv      = (op_code / info.li_line_range)
                        * info.li_min_insn_length;
          advAddr = adv;
          state_machine_regs.address += adv;
@@ -800,7 +827,7 @@ void read_dwarf2_lineblock ( struct _DebugInfo* di,
             break;
 
          case DW_LNS_advance_pc:
-            adv = info.li_min_insn_length 
+            adv = info.li_min_insn_length
                      * read_leb128 (data, & bytes_read, 0);
             data += bytes_read;
             state_machine_regs.address += adv;
@@ -979,7 +1006,7 @@ void read_unitinfo_dwarf2( /*OUT*/UnitInfo* ui,
    blklen = read_initial_length_field( p, &ui->dw64 );
    p += ui->dw64 ? 12 : 4;
 
-   /* version should be 2 */
+   /* version should be 2, 3 or 4 */
    ver = *((UShort*)p);
    p += 2;
 
@@ -1049,6 +1076,9 @@ void read_unitinfo_dwarf2( /*OUT*/UnitInfo* ui,
             classes) use FORM_data8, not FORM_data4.  Also,
             FORM_ref_addr and FORM_strp are 64-bit values, not 32-bit
             values. */
+         /* TJH 27 Apr 10: in DWARF 4 lineptr (and loclistptr,macptr,
+            rangelistptr classes) use FORM_sec_offset which is 64 bits
+            in 64 bit DWARF and 32 bits in 32 bit DWARF. */
          switch( form ) {
             /* Those cases extract the data properly */
             case 0x05: /* FORM_data2 */     cval = *((UShort*)p); p +=2; break;
@@ -1066,7 +1096,11 @@ void read_unitinfo_dwarf2( /*OUT*/UnitInfo* ui,
             case 0x08: /* FORM_string */    sval = (Char*)p; 
                                             p += VG_(strlen)((Char*)p) + 1; break;
             case 0x0b: /* FORM_data1 */     cval = *p; p++; break;
-            
+            case 0x17: /* FORM_sec_offset */if (ui->dw64) {
+                                               cval = *((ULong*)p); p += 8;
+                                            } else {
+                                               cval = *((UInt*)p); p += 4;
+                                            }; break;
             /* TODO : Following ones just skip data - implement if you need */
             case 0x01: /* FORM_addr */      p += addr_size; break;
             case 0x03: /* FORM_block2 */    p += *((UShort*)p) + 2; break;
@@ -1085,7 +1119,10 @@ void read_unitinfo_dwarf2( /*OUT*/UnitInfo* ui,
             case 0x13: /* FORM_ref4 */      p += 4; break;
             case 0x14: /* FORM_ref8 */      p += 8; break;
             case 0x15: /* FORM_ref_udata */ read_leb128U( &p ); break;
-            
+            case 0x18: /* FORM_exprloc */   p += read_leb128U( &p ); break;
+            case 0x19: /* FORM_flag_present */break;
+            case 0x20: /* FORM_ref_sig8 */  p += 8; break;
+
             default:
                VG_(printf)( "### unhandled dwarf2 abbrev form code 0x%x\n", form );
                break;
@@ -1163,9 +1200,9 @@ void ML_(read_debuginfo_dwarf3)
 
       /* version should be 2 */
       ver = *((UShort*)( block_img + blklen_len ));
-      if ( ver != 2 && ver != 3 ) {
+      if ( ver != 2 && ver != 3 && ver != 4 ) {
          ML_(symerr)( di, True,
-                      "Ignoring non-Dwarf2/3 block in .debug_info" );
+                      "Ignoring non-Dwarf2/3/4 block in .debug_info" );
          continue;
       }
       
@@ -3705,8 +3742,8 @@ void ML_(read_callframe_info_dwarf3)
             VG_(printf)("cie.version     = %d\n", (Int)cie_version);
          if (di->ddump_frames)
             VG_(printf)("  Version:               %d\n", (Int)cie_version);
-         if (cie_version != 1 && cie_version != 3) {
-            how = "unexpected CIE version (not 1 nor 3)";
+         if (cie_version != 1 && cie_version != 3 && cie_version != 4) {
+            how = "unexpected CIE version (not 1 nor 3 nor 4)";
             goto bad;
          }
 
@@ -3722,6 +3759,19 @@ void ML_(read_callframe_info_dwarf3)
             cie_augmentation += 2;
          }
 
+         if (cie_version >= 4) {
+            if (read_UChar(data) != sizeof(Addr)) {
+               how = "unexpected address size";
+               goto bad;
+            }
+            data += sizeof(UChar);
+            if (read_UChar(data) != 0) {
+               how = "unexpected non-zero segment size";
+               goto bad;
+            }
+            data += sizeof(UChar);
+         }
+
          the_CIEs[this_CIE].code_a_f = read_leb128( data, &nbytes, 0);
          data += nbytes;
          if (di->trace_cfi) 
diff --git a/coregrind/m_debuginfo/readdwarf3.c b/coregrind/m_debuginfo/readdwarf3.c
index e9ec841d..45263200 100644
--- a/coregrind/m_debuginfo/readdwarf3.c
+++ b/coregrind/m_debuginfo/readdwarf3.c
@@ -1,6 +1,6 @@
 
 /*--------------------------------------------------------------------*/
-/*--- Read DWARF3 ".debug_info" sections (DIE trees).              ---*/
+/*--- Read DWARF3/4 ".debug_info" sections (DIE trees).            ---*/
 /*---                                                 readdwarf3.c ---*/
 /*--------------------------------------------------------------------*/
 
@@ -387,7 +387,7 @@ typedef
       void (*barf)( HChar* ) __attribute__((noreturn));
       /* Is this 64-bit DWARF ? */
       Bool   is_dw64;
-      /* Which DWARF version ?  (2 or 3) */
+      /* Which DWARF version ?  (2, 3 or 4) */
       UShort version;
       /* Length of this Compilation Unit, as stated in the
          .unit_length :: InitialLength field of the CU Header.
@@ -805,8 +805,8 @@ void parse_CU_Header ( /*OUT*/CUConst* cc,
 
    /* version */
    cc->version = get_UShort( c );
-   if (cc->version != 2 && cc->version != 3)
-      cc->barf( "parse_CU_Header: is neither DWARF2 nor DWARF3" );
+   if (cc->version != 2 && cc->version != 3 && cc->version != 4)
+      cc->barf( "parse_CU_Header: is neither DWARF2 nor DWARF3 nor DWARF4" );
    TRACE_D3("   Version:       %d\n", (Int)cc->version );
 
    /* debug_abbrev_offset */
@@ -984,11 +984,21 @@ void get_Form_contents ( /*OUT*/ULong* cts,
          *ctsSzB = 8;
          TRACE_D3("%llu", *cts);
          break;
+      case DW_FORM_sec_offset:
+         *cts = (ULong)get_Dwarfish_UWord( c, cc->is_dw64 );
+         *ctsSzB = cc->is_dw64 ? 8 : 4;
+         TRACE_D3("%llu", *cts);
+         break;
       case DW_FORM_sdata:
          *cts = (ULong)(Long)get_SLEB128(c);
          *ctsSzB = 8;
          TRACE_D3("%lld", (Long)*cts);
          break;
+      case DW_FORM_udata:
+         *cts = (ULong)(Long)get_ULEB128(c);
+         *ctsSzB = 8;
+         TRACE_D3("%llu", (Long)*cts);
+         break;
       case DW_FORM_addr:
          /* note, this is a hack.  DW_FORM_addr is defined as getting
             a word the size of the target machine as defined by the
@@ -1055,6 +1065,22 @@ void get_Form_contents ( /*OUT*/ULong* cts,
          *ctsMemSzB = 1 + (ULong)VG_(strlen)(str);
          break;
       }
+      case DW_FORM_ref1: {
+         UChar  u8 = get_UChar(c);
+         UWord res = cc->cu_start_offset + (UWord)u8;
+         *cts = (ULong)res;
+         *ctsSzB = sizeof(UWord);
+         TRACE_D3("<%lx>", res);
+         break;
+      }
+      case DW_FORM_ref2: {
+         UShort  u16 = get_UShort(c);
+         UWord res = cc->cu_start_offset + (UWord)u16;
+         *cts = (ULong)res;
+         *ctsSzB = sizeof(UWord);
+         TRACE_D3("<%lx>", res);
+         break;
+      }
       case DW_FORM_ref4: {
          UInt  u32 = get_UInt(c);
          UWord res = cc->cu_start_offset + (UWord)u32;
@@ -1063,6 +1089,22 @@ void get_Form_contents ( /*OUT*/ULong* cts,
          TRACE_D3("<%lx>", res);
          break;
       }
+      case DW_FORM_ref8: {
+         ULong  u64 = get_ULong(c);
+         UWord res = cc->cu_start_offset + (UWord)u64;
+         *cts = (ULong)res;
+         *ctsSzB = sizeof(UWord);
+         TRACE_D3("<%lx>", res);
+         break;
+      }
+      case DW_FORM_ref_udata: {
+         ULong  u64 = get_ULEB128(c);
+         UWord res = cc->cu_start_offset + (UWord)u64;
+         *cts = (ULong)res;
+         *ctsSzB = sizeof(UWord);
+         TRACE_D3("<%lx>", res);
+         break;
+      }
       case DW_FORM_flag: {
          UChar u8 = get_UChar(c);
          TRACE_D3("%u", (UInt)u8);
@@ -1070,6 +1112,11 @@ void get_Form_contents ( /*OUT*/ULong* cts,
          *ctsSzB = 1;
          break;
       }
+      case DW_FORM_flag_present:
+         TRACE_D3("1");
+         *cts = 1;
+         *ctsSzB = 1;
+         break;
       case DW_FORM_block1: {
          ULong  u64b;
          ULong  u64 = (ULong)get_UChar(c);
@@ -1096,6 +1143,50 @@ void get_Form_contents ( /*OUT*/ULong* cts,
          *ctsMemSzB = (UWord)u64;
          break;
       }
+      case DW_FORM_block4: {
+         ULong  u64b;
+         ULong  u64 = (ULong)get_UInt(c);
+         UChar* block = get_address_of_Cursor(c);
+         TRACE_D3("%llu byte block: ", u64);
+         for (u64b = u64; u64b > 0; u64b--) {
+            UChar u8 = get_UChar(c);
+            TRACE_D3("%x ", (UInt)u8);
+         }
+         *cts = (ULong)(UWord)block;
+         *ctsMemSzB = (UWord)u64;
+         break;
+      }
+      case DW_FORM_exprloc:
+      case DW_FORM_block: {
+         ULong  u64b;
+         ULong  u64 = (ULong)get_ULEB128(c);
+         UChar* block = get_address_of_Cursor(c);
+         TRACE_D3("%llu byte block: ", u64);
+         for (u64b = u64; u64b > 0; u64b--) {
+            UChar u8 = get_UChar(c);
+            TRACE_D3("%x ", (UInt)u8);
+         }
+         *cts = (ULong)(UWord)block;
+         *ctsMemSzB = (UWord)u64;
+         break;
+      }
+      case DW_FORM_ref_sig8: {
+         ULong  u64b;
+         UChar* block = get_address_of_Cursor(c);
+         TRACE_D3("8 byte signature: ");
+         for (u64b = 8; u64b > 0; u64b--) {
+            UChar u8 = get_UChar(c);
+            TRACE_D3("%x ", (UInt)u8);
+         }
+         *cts = (ULong)(UWord)block;
+         *ctsMemSzB = 8;
+         break;
+      }
+      case DW_FORM_indirect:
+         get_Form_contents (cts, ctsSzB, ctsMemSzB, cc, c, td3,
+                            (DW_FORM)get_ULEB128(c));
+         return;
+
       default:
          VG_(printf)(
             "get_Form_contents: unhandled %d (%s) at <%lx>\n",
@@ -1322,11 +1413,13 @@ void read_filename_table( /*MOD*/D3VarParser* parser,
       get_Initial_Length( &is_dw64, &c,
            "read_filename_table: invalid initial-length field" );
    version = get_UShort( &c );
-   if (version != 2 && version != 3)
-     cc->barf("read_filename_table: Only DWARF version 2 and 3 line info "
+   if (version != 2 && version != 3 && version != 4)
+     cc->barf("read_filename_table: Only DWARF version 2, 3 and 4 line info "
               "is currently supported.");
    /*header_length              = (ULong)*/ get_Dwarfish_UWord( &c, is_dw64 );
    /*minimum_instruction_length = */ get_UChar( &c );
+   if (version >= 4)
+      /*maximum_operations_per_insn = */ get_UChar( &c );
    /*default_is_stmt            = */ get_UChar( &c );
    /*line_base                  = (Char)*/ get_UChar( &c );
    /*line_range                 = */ get_UChar( &c );
@@ -2025,7 +2118,7 @@ static void parse_type_DIE ( /*MOD*/XArray* /* of TyEnt */ tyents,
             case DW_LANG_C89: case DW_LANG_C:
             case DW_LANG_C_plus_plus: case DW_LANG_ObjC:
             case DW_LANG_ObjC_plus_plus: case DW_LANG_UPC:
-            case DW_LANG_Upc:
+            case DW_LANG_Upc: case DW_LANG_C99:
                parser->language = 'C'; break;
             case DW_LANG_Fortran77: case DW_LANG_Fortran90:
             case DW_LANG_Fortran95:
@@ -2033,8 +2126,8 @@ static void parse_type_DIE ( /*MOD*/XArray* /* of TyEnt */ tyents,
             case DW_LANG_Ada83: case DW_LANG_Cobol74:
             case DW_LANG_Cobol85: case DW_LANG_Pascal83:
             case DW_LANG_Modula2: case DW_LANG_Java:
-            case DW_LANG_C99: case DW_LANG_Ada95:
-            case DW_LANG_PLI: case DW_LANG_D:
+            case DW_LANG_Ada95: case DW_LANG_PLI:
+            case DW_LANG_D: case DW_LANG_Python:
             case DW_LANG_Mips_Assembler:
                parser->language = '?'; break;
             default: