ed/raddebugger
1
0
Fork 0
mirror of https://github.com/Ed94/raddebugger.git synced 2026-07-09 19:11:38 -07:00
Code Issues Packages Projects Releases Wiki Activity

raddbgi_from_pdb: parameterize symbol stream conversion pass with sub-range of symbols; this cannot be applied to unit conversion, since they have cross-sym-record concepts, but it *can* be used for global symbol streams, which are also the most expensive to convert. this majorly speeds up global symbol stream conversion

Browse Source
This commit is contained in:
Ryan Fleury
2024-02-19 16:41:22 -08:00
parent 0b73582b87
commit 10a0420264
2 changed files with 52 additions and 40 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/raddbgi_from_pdb/raddbgi_from_pdb.c
+34 -24
View File
@@ -710,10 +710,10 @@ p2r_itype_fwd_map_fill_task_thread__entry_point(void *p)
}
////////////////////////////////
//~ rjf: Per-Unit Symbol Conversion Pass Thread Entry Point
//~ rjf: Symbol Stream Conversion Path & Thread
internal P2R_UnitSymbolConvertOut *
p2r_unit_symbol_convert(Arena *arena, P2R_UnitSymbolConvertIn *in)
internal P2R_SymbolStreamConvertOut *
p2r_symbol_stream_convert(Arena *arena, P2R_SymbolStreamConvertIn *in)
{
Temp scratch = scratch_begin(&arena, 1);
#define p2r_type_ptr_from_itype(itype) (((itype) < in->tpi_leaf->itype_opl) ? (in->itype_type_ptrs[(in->itype_fwd_map[(itype)] ? in->itype_fwd_map[(itype)] : (itype))]) : 0)
@@ -734,13 +734,13 @@ p2r_unit_symbol_convert(Arena *arena, P2R_UnitSymbolConvertIn *in)
//- rjf: symbols pass 1: produce procedure frame info map (procedure -> frame info)
//
U64 procedure_frameprocs_count = 0;
U64 procedure_frameprocs_cap = in->sym->sym_ranges.count;
U64 procedure_frameprocs_cap = (in->sym_ranges_opl - in->sym_ranges_first);
CV_SymFrameproc **procedure_frameprocs = push_array_no_zero(scratch.arena, CV_SymFrameproc *, procedure_frameprocs_cap);
ProfScope("symbols pass 1: produce procedure frame info map (procedure -> frame info)")
{
U64 procedure_num = 0;
CV_RecRange *rec_ranges_first = in->sym->sym_ranges.ranges;
CV_RecRange *rec_ranges_opl = rec_ranges_first + in->sym->sym_ranges.count;
CV_RecRange *rec_ranges_first = in->sym->sym_ranges.ranges + in->sym_ranges_first;
CV_RecRange *rec_ranges_opl = in->sym->sym_ranges.ranges + in->sym_ranges_opl;
for(CV_RecRange *rec_range = rec_ranges_first;
rec_range < rec_ranges_opl;
rec_range += 1)
@@ -801,8 +801,8 @@ p2r_unit_symbol_convert(Arena *arena, P2R_UnitSymbolConvertIn *in)
RDIM_LocationSet *defrange_target = 0;
B32 defrange_target_is_param = 0;
U64 procedure_num = 0;
CV_RecRange *rec_ranges_first = in->sym->sym_ranges.ranges;
CV_RecRange *rec_ranges_opl = rec_ranges_first + in->sym->sym_ranges.count;
CV_RecRange *rec_ranges_first = in->sym->sym_ranges.ranges + in->sym_ranges_first;
CV_RecRange *rec_ranges_opl = in->sym->sym_ranges.ranges + in->sym_ranges_opl;
typedef struct P2R_ScopeNode P2R_ScopeNode;
struct P2R_ScopeNode
{
@@ -1407,7 +1407,7 @@ p2r_unit_symbol_convert(Arena *arena, P2R_UnitSymbolConvertIn *in)
//////////////////////////
//- rjf: allocate & fill output
//
P2R_UnitSymbolConvertOut *out = push_array(arena, P2R_UnitSymbolConvertOut, 1);
P2R_SymbolStreamConvertOut *out = push_array(arena, P2R_SymbolStreamConvertOut, 1);
{
out->procedures = sym_procedures;
out->global_variables = sym_global_variables;
@@ -1421,9 +1421,9 @@ p2r_unit_symbol_convert(Arena *arena, P2R_UnitSymbolConvertIn *in)
}
internal void
p2r_unit_symbol_convert_task_thread__entry_point(void *p)
p2r_symbol_stream_convert_task_thread__entry_point(void *p)
{
P2R_UnitSymbolTaskBatch *batch = (P2R_UnitSymbolTaskBatch *)p;
P2R_SymbolStreamTaskBatch *batch = (P2R_SymbolStreamTaskBatch *)p;
ThreadName("[p2r] unit symbol thread");
for(;;)
{
@@ -1432,10 +1432,10 @@ p2r_unit_symbol_convert_task_thread__entry_point(void *p)
{
break;
}
P2R_UnitSymbolTask *task = &batch->tasks[next_task_num-1];
P2R_SymbolStreamTask *task = &batch->tasks[next_task_num-1];
ProfScope("convert (task #%I64u)", next_task_num)
{
task->convert_out = p2r_unit_symbol_convert(task->out_arena, &task->convert_in);
task->convert_out = p2r_symbol_stream_convert(task->out_arena, &task->convert_in);
}
}
}
@@ -3291,21 +3291,22 @@ p2r_convert(Arena *arena, P2R_ConvertIn *in)
}
//////////////////////////////////////////////////////////////
//- rjf: produce symbols from all sym streams
//- rjf: produce symbols from all streams
//
RDIM_SymbolChunkList all_procedures = {0};
RDIM_SymbolChunkList all_global_variables = {0};
RDIM_SymbolChunkList all_thread_variables = {0};
RDIM_ScopeChunkList all_scopes = {0};
ProfScope("produce symbols from all sym streams")
ProfScope("produce symbols from all streams")
{
////////////////////////////
//- rjf: produce per-task info
//
U64 tasks_count = 1+comp_unit_count;
P2R_UnitSymbolTask *tasks = push_array(scratch.arena, P2R_UnitSymbolTask, tasks_count);
U64 global_stream_subdivision_tasks_count = (sym->sym_ranges.count+65535)/65536;
U64 global_stream_syms_per_task = sym->sym_ranges.count/global_stream_subdivision_tasks_count;
U64 tasks_count = comp_unit_count + global_stream_subdivision_tasks_count;
P2R_SymbolStreamTask *tasks = push_array(scratch.arena, P2R_SymbolStreamTask, tasks_count);
{
tasks[0].convert_in.sym = sym;
for(U64 idx = 0; idx < tasks_count; idx += 1)
{
tasks[idx].convert_in.arch = arch;
@@ -3316,9 +3317,18 @@ p2r_convert(Arena *arena, P2R_ConvertIn *in)
tasks[idx].convert_in.itype_type_ptrs = itype_type_ptrs;
tasks[idx].convert_in.link_name_map = &link_name_map;
tasks[idx].out_arena = arena_alloc();
if(idx > 0)
if(idx < global_stream_subdivision_tasks_count)
{
tasks[idx].convert_in.sym = sym_for_unit[idx-1];
tasks[idx].convert_in.sym = sym;
tasks[idx].convert_in.sym_ranges_first= idx*global_stream_syms_per_task;
tasks[idx].convert_in.sym_ranges_opl = tasks[idx].convert_in.sym_ranges_first + global_stream_syms_per_task;
tasks[idx].convert_in.sym_ranges_opl = ClampTop(tasks[idx].convert_in.sym_ranges_opl, sym->sym_ranges.count);
}
else
{
tasks[idx].convert_in.sym = sym_for_unit[idx-global_stream_subdivision_tasks_count];
tasks[idx].convert_in.sym_ranges_first= 0;
tasks[idx].convert_in.sym_ranges_opl = sym_for_unit[idx-global_stream_subdivision_tasks_count]->sym_ranges.count;
}
}
}
@@ -3327,7 +3337,7 @@ p2r_convert(Arena *arena, P2R_ConvertIn *in)
//- rjf: produce task batch info
//
U64 num_tasks_taken = 0;
P2R_UnitSymbolTaskBatch task_batch = {0};
P2R_SymbolStreamTaskBatch task_batch = {0};
{
task_batch.tasks = tasks;
task_batch.tasks_count = tasks_count;
@@ -3341,7 +3351,7 @@ p2r_convert(Arena *arena, P2R_ConvertIn *in)
OS_Handle *task_threads = push_array(scratch.arena, OS_Handle, task_threads_count);
for(U64 idx = 0; idx < task_threads_count; idx += 1)
{
task_threads[idx] = os_launch_thread(p2r_unit_symbol_convert_task_thread__entry_point, &task_batch, 0);
task_threads[idx] = os_launch_thread(p2r_symbol_stream_convert_task_thread__entry_point, &task_batch, 0);
}
////////////////////////////
@@ -3356,10 +3366,10 @@ p2r_convert(Arena *arena, P2R_ConvertIn *in)
{
break;
}
P2R_UnitSymbolTask *task = &task_batch.tasks[next_task_num-1];
P2R_SymbolStreamTask *task = &task_batch.tasks[next_task_num-1];
ProfScope("convert (task #%I64u)", next_task_num)
{
task->convert_out = p2r_unit_symbol_convert(task->out_arena, &task->convert_in);
task->convert_out = p2r_symbol_stream_convert(task->out_arena, &task->convert_in);
}
}
}
src/raddbgi_from_pdb/raddbgi_from_pdb.h
+18 -16
View File
@@ -155,23 +155,25 @@ struct P2R_ITypeFwdMapFillTaskBatch
U64 *num_tasks_taken_ptr;
};
//- rjf: per-unit symbol conversion
//- rjf: symbol stream conversion
typedef struct P2R_UnitSymbolConvertIn P2R_UnitSymbolConvertIn;
struct P2R_UnitSymbolConvertIn
typedef struct P2R_SymbolStreamConvertIn P2R_SymbolStreamConvertIn;
struct P2R_SymbolStreamConvertIn
{
RDI_Arch arch;
PDB_CoffSectionArray *coff_sections;
PDB_TpiHashParsed *tpi_hash;
CV_LeafParsed *tpi_leaf;
CV_SymParsed *sym;
U64 sym_ranges_first;
U64 sym_ranges_opl;
CV_TypeId *itype_fwd_map;
RDIM_Type **itype_type_ptrs;
P2R_LinkNameMap *link_name_map;
};
typedef struct P2R_UnitSymbolConvertOut P2R_UnitSymbolConvertOut;
struct P2R_UnitSymbolConvertOut
typedef struct P2R_SymbolStreamConvertOut P2R_SymbolStreamConvertOut;
struct P2R_SymbolStreamConvertOut
{
RDIM_SymbolChunkList procedures;
RDIM_SymbolChunkList global_variables;
@@ -179,21 +181,21 @@ struct P2R_UnitSymbolConvertOut
RDIM_ScopeChunkList scopes;
};
typedef struct P2R_UnitSymbolTask P2R_UnitSymbolTask;
struct P2R_UnitSymbolTask
typedef struct P2R_SymbolStreamTask P2R_SymbolStreamTask;
struct P2R_SymbolStreamTask
{
// rjf: inputs
P2R_UnitSymbolConvertIn convert_in;
P2R_SymbolStreamConvertIn convert_in;
// rjf: outputs
Arena *out_arena;
P2R_UnitSymbolConvertOut *convert_out;
P2R_SymbolStreamConvertOut *convert_out;
};
typedef struct P2R_UnitSymbolTaskBatch P2R_UnitSymbolTaskBatch;
struct P2R_UnitSymbolTaskBatch
typedef struct P2R_SymbolStreamTaskBatch P2R_SymbolStreamTaskBatch;
struct P2R_SymbolStreamTaskBatch
{
P2R_UnitSymbolTask *tasks;
P2R_SymbolStreamTask *tasks;
U64 tasks_count;
U64 *num_tasks_taken_ptr;
};
@@ -238,16 +240,16 @@ internal void p2r_tpi_leaf_parse_thread__entry_point(void *p);
internal void p2r_exe_hash_thread__entry_point(void *p);
////////////////////////////////
//~ rjf: Type Forward Resolution Map Build / Thread
//~ rjf: Type Forward Resolution Map Build Path & Thread
internal void p2r_itype_fwd_map_fill(P2R_ITypeFwdMapFillIn *in);
internal void p2r_itype_fwd_map_fill_task_thread__entry_point(void *p);
////////////////////////////////
//~ rjf: Per-Unit Symbol Conversion / Thread
//~ rjf: Symbol Stream Conversion Paths & Thread
internal P2R_UnitSymbolConvertOut *p2r_unit_symbol_convert(Arena *arena, P2R_UnitSymbolConvertIn *in);
internal void p2r_unit_symbol_convert_task_thread__entry_point(void *p);
internal P2R_SymbolStreamConvertOut *p2r_symbol_stream_convert(Arena *arena, P2R_SymbolStreamConvertIn *in);
internal void p2r_symbol_stream_convert_task_thread__entry_point(void *p);
////////////////////////////////
//~ rjf: Top-Level Conversion Entry Point