many fixes

* fix compile warnning and errors
* fix bugs in CMakeLists
* remove xxhash from source and use system-level xxhash so that macos can compile
* change global variable names
* add n_core function for macos

CMakeLists.txt

@@ -32,10 +32,10 @@ set_property(CACHE LOG_LEVEL PROPERTY STRINGS INFO WARN ERROR DEBUG VERBOSE VVER
 # detect platform #
 ########################################
 if (${CMAKE_SYSTEM_NAME} MATCHES "Darwin")
+    message(STATUS "Mac OS X detected, version ${CMAKE_SYSTEM_VERSION}")
     add_definitions(-DOS_DARWIN)
-    if (NOT CMAKE_BUILD_TYPE)
-        set(CMAKE_BUILD_TYPE Debug)
-    endif()
+    set(CMAKE_INCLUDE_PATH ${CMAKE_INCLUDE_PATH} "/opt/homebrew/include/")
+    set(CMAKE_PREFIX_PATH ${CMAKE_PREFIX_PATH} "/opt/homebrew/")
 
 elseif (${CMAKE_SYSTEM_NAME} MATCHES "Linux")
 #    set(CFLAGS "$ENV{CFLAGS} " "-Wl,--export-dynamic ")
@@ -121,7 +121,6 @@ set(CFLAGS "$ENV{CFLAGS} "
         # "-Wmissing-prototypes "
         "-Wmissing-declarations "
         "-Wredundant-decls "
-        "-Wunused-value -Wunused-variable "
         "-fno-strict-aliasing "
         )
 
@@ -142,21 +141,22 @@ set(CMAKE_MODULE_PATH ${CMAKE_MODULE_PATH} "${CMAKE_SOURCE_DIR}/cmake/Modules/")
 find_package(GLib REQUIRED)
 include_directories(${GLib_INCLUDE_DIRS})
 set(LIBS ${LIBS} ${GLib_LIBRARY})
-#if (${CMAKE_SYSTEM_NAME} MATCHES "Darwin")
-#    pkg_check_modules(GLib2 REQUIRED Glib-2.0)
-#    include_directories(${GLib2_INCLUDE_DIRS})
-#    #    add_compile_options("${LDFLAGS}")
-#    link_libraries(${GLib2_LDFLAGS})
-#endif()
+
+
+find_package(argp REQUIRED)
+include_directories(${ARGP_INCLUDE_DIRS})
+set(LIBS ${LIBS} ${ARGP_LIBRARY})
 
 if (OPT_SUPPORT_ZSTD_TRACE)
     add_compile_definitions(SUPPORT_ZSTD_TRACE=1)
     find_package(ZSTD)
+    # https://stackoverflow.com/questions/61377055/cannot-find-gflags-gflags-h-while-building-library-osx/61379123#61379123
+    include_directories(${ZSTD_INCLUDE_DIR})
     if ("${ZSTD_LIBRARIES}" STREQUAL "") 
         message(FATAL_ERROR "zstd not found")
     endif()
     link_libraries(${ZSTD_LIBRARIES})
-    message(STATUS "ZSTD_INCLUDE_DIRS ${ZSTD_INCLUDE_DIRS}, ZSTD_LIBRARIES ${ZSTD_LIBRARIES}")
+    message(STATUS "ZSTD_INCLUDE_DIR ${ZSTD_INCLUDE_DIR}, ZSTD_LIBRARIES ${ZSTD_LIBRARIES}")
 else()
     remove_definitions(SUPPORT_ZSTD_TRACE)
 endif(OPT_SUPPORT_ZSTD_TRACE)
@@ -224,7 +224,7 @@ message(STATUS "<<++=====------------------/\\------------------=====++>>")
 message(STATUS "================== dependency related ==================")
 message(STATUS "glib found?     ${GLib_FOUND} - LIBS=${GLib_LIBRARY}, header =${GLib_INCLUDE_DIRS}")
 message(STATUS "tcmalloc found? ${Tcmalloc_FOUND} - LIBS=${Tcmalloc_LIBRARIES}, header=${Tcmalloc_INCLUDE_DIRS}")
-message(STATUS "ZSTD found? ${ZSTD_FOUND} - LIBS=${ZSTD_LIBRARIES}, header=${ZSTD_INCLUDE_DIRS}")
+message(STATUS "ZSTD found? ${ZSTD_FOUND} - LIBS=${ZSTD_LIBRARIES}, header=${ZSTD_INCLUDE_DIR}")
 
 message(STATUS "==================== CMake related =====================")
 message(STATUS "platform          = ${CMAKE_SYSTEM_NAME} ${CMAKE_SYSTEM_VERSION}")
@@ -330,7 +330,7 @@ endif(OPT_SUPPORT_ZSTD_TRACE)
 file(GLOB dataStructure_source
         ${PROJECT_SOURCE_DIR}/libCacheSim/dataStructure/*.c
         ${PROJECT_SOURCE_DIR}/libCacheSim/dataStructure/hashtable/*.c
-        ${PROJECT_SOURCE_DIR}/libCacheSim/dataStructure/hash/*.c
+        ${PROJECT_SOURCE_DIR}/libCacheSim/dataStructure/hash/murmur3.c
     )
 
 file(GLOB profiler_source
@@ -347,7 +347,8 @@ set(LIB_SOURCE ${LIB_SOURCE} ${cache_source} ${reader_source} ${dataStructure_so
 # # https://stackoverflow.com/questions/32469953/why-is-cmake-designed-so-that-it-removes-runtime-path-when-installing
 SET(CMAKE_INSTALL_RPATH_USE_LINK_PATH TRUE)
 
-add_library(${PROJECT_NAME} SHARED ${LIB_SOURCE})
+add_library(${PROJECT_NAME} ${LIB_SOURCE})
+# add_library(${PROJECT_NAME} SHARED ${LIB_SOURCE})
 set_target_properties(${PROJECT_NAME} PROPERTIES VERSION ${${PROJECT_NAME}_VERSION})
 set_target_properties(${PROJECT_NAME} PROPERTIES LINKER_LANGUAGE C)
 set_target_properties(${PROJECT_NAME} PROPERTIES PUBLIC_HEADER ${PROJECT_SOURCE_DIR}/libCacheSim/include/libCacheSim.h)
@@ -366,12 +367,6 @@ install(TARGETS ${PROJECT_NAME}
     LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR}
     PUBLIC_HEADER DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
 
-
-
-# #target_compile_definitions(binBenchmark PRIVATE HASH=0xb2)
-# #add_executable(binBenchmark libCacheSim/benchmark/main.c)
-# #target_link_libraries(binBenchmark benchmark cachelib evictionAlgos traceReader profiler dataStructure utils
-# #        ${LIBS})
 #############################
 
 

cmake/Modules/Findargp.cmake

@@ -0,0 +1,84 @@
+# - try to find the argp library/component of glibc
+#
+# Users may optionally supply:
+#  ARGP_ROOT_DIR - a prefix to start searching.
+#
+# Cache Variables: (probably not for direct use in your scripts)
+#  ARGP_INCLUDE_DIR
+#  ARGP_LIBRARY, only defined if linking to an extra library is required
+#
+# Non-cache variables you might use in your CMakeLists.txt:
+#  ARGP_FOUND
+#  ARGP_INCLUDE_DIRS
+#  ARGP_LIBRARIES
+#
+# Requires these CMake modules:
+#  FindPackageHandleStandardArgs (known included with CMake >=2.6.2)
+#
+# Original Author:
+# 2009-2010 Rylie Pavlik <rylie@ryliepavlik.com>
+# https://ryliepavlik.com/
+# Iowa State University HCI Graduate Program/VRAC
+#
+# Copyright 2009-2010, Iowa State University
+#
+# Distributed under the Boost Software License, Version 1.0.
+# (See accompanying file LICENSE_1_0.txt or copy at
+# http://www.boost.org/LICENSE_1_0.txt)
+#
+# SPDX-License-Identifier: BSL-1.0
+
+set(ARGP_ROOT_DIR
+	"${ARGP_ROOT_DIR}"
+	CACHE
+	PATH
+	"Path to search for ARGP library")
+
+###
+# Configure ARGP
+###
+set(_check ARGP_INCLUDE_DIR)
+
+find_path(ARGP_INCLUDE_DIR
+	NAMES
+	argp.h
+	HINTS
+	"${ARGP_ROOT_DIR}"
+	PATHS
+	/usr/local
+	/opt/local
+	/sw)
+mark_as_advanced(ARGP_INCLUDE_DIR)
+
+include(CheckFunctionExists)
+check_function_exists(argp_parse ARGP_BUILTIN)
+
+if(NOT ARGP_BUILTIN)
+	find_library(ARGP_LIBRARY
+		NAMES
+		argp
+		HINTS
+		"${ARGP_ROOT_DIR}"
+		PATH_SUFFIXES
+		lib
+		lib64
+		PATHS
+		/usr/local
+		/opt/local
+		/sw)
+	list(APPEND _check ARGP_LIBRARY)
+	mark_as_advanced(ARGP_LIBRARY)
+endif()
+
+# handle the QUIETLY and REQUIRED arguments and set xxx_FOUND to TRUE if
+# all listed variables are TRUE
+include(FindPackageHandleStandardArgs)
+find_package_handle_standard_args(argp
+	DEFAULT_MSG
+	${_check})
+
+if(ARGP_FOUND)
+	set(ARGP_INCLUDE_DIRS "${ARGP_INCLUDE_DIR}")
+	set(ARGP_LIBRARIES "${ARGP_LIBRARY}")
+	mark_as_advanced(ARGP_ROOT_DIR)
+endif()

libCacheSim/bin/cachesim/cli_parser.c

@@ -477,7 +477,7 @@ static void set_cache_size(struct arguments *args, reader_t *reader) {
   args->n_cache_size = n_cache_sizes;
 
   if (args->n_cache_size == 0) {
-    printf("working set %ld too small\n", wss);
+    printf("working set %ld too small\n", (long) wss);
     exit(0);
   }
 }
@@ -494,7 +494,7 @@ void print_parsed_args(struct arguments *args) {
       output_str, OUTPUT_STR_LEN - 1,
       "trace path: %s, trace_type %s, ofilepath "
       "%s, %d threads, warmup %d sec, total %d algo x %d size = %d caches",
-      args->trace_path, trace_type_str[args->trace_type], args->ofilepath,
+      args->trace_path, g_trace_type_name[args->trace_type], args->ofilepath,
       args->n_thread, args->warmup_sec, args->n_eviction_algo,
       args->n_cache_size, args->n_eviction_algo * args->n_cache_size);
 

libCacheSim/bin/cli_reader_utils.c

@@ -1,13 +1,12 @@
 
 
 #define _GNU_SOURCE
-#include "cli_reader_utils.h"
-
 #include <assert.h>
 #include <string.h>
 
 #include "../include/libCacheSim/reader.h"
 #include "../utils/include/mystr.h"
+#include "cli_reader_utils.h"
 
 #ifdef __cplusplus
 extern "C" {
@@ -60,11 +59,12 @@ trace_type_e trace_type_str_to_enum(const char *trace_type_str,
     return ORACLE_CF1_TRACE;
   } else if (strcasecmp(trace_type_str, "oracleSysTwrNS") == 0) {
     return ORACLE_SYS_TWRNS_TRACE;
-  } else if (strcasecmp(trace_type_str, "valpinTrace") == 0){
+  } else if (strcasecmp(trace_type_str, "valpinTrace") == 0) {
     return VALPIN_TRACE;
   } else {
     ERROR("unsupported trace type: %s\n", trace_type_str);
   }
+  return UNKNOWN_TRACE;
 }
 
 bool is_true(const char *arg) {
@@ -215,7 +215,7 @@ trace_type_e detect_trace_type(const char *trace_path) {
     trace_type = UNKNOWN_TRACE;
   }
 
-  INFO("detecting trace type: %s\n", trace_type_str[trace_type]);
+  INFO("detecting trace type: %s\n", g_trace_type_name[trace_type]);
   return trace_type;
 }
 
@@ -276,7 +276,8 @@ void cal_working_set_size(reader_t *reader, int64_t *wss_obj,
   }
   *wss_obj *= scaling_factor;
   *wss_byte *= scaling_factor;
-  INFO("working set size: %ld object %ld byte\n", *wss_obj, *wss_byte);
+  INFO("working set size: %ld object %ld byte\n", (long)*wss_obj,
+       (long)*wss_byte);
 
   free_request(req);
   reset_reader(reader);

libCacheSim/bin/traceAnalyzer/cli_parser.cpp

@@ -61,9 +61,9 @@ enum argp_option_short {
 static struct argp_option options[] = {
     {NULL, 0, NULL, 0, "trace reader related parameters:", 0},
     {"trace-type-params", OPTION_TRACE_TYPE_PARAMS,
-     "time-col=1,obj-id-col=2,obj-size-col=3,delimiter=,", NULL,
+     "time-col=1,obj-id-col=2,obj-size-col=3,delimiter=,", 0,
      "Parameters used for csv trace", 1},
-    {"num-req", OPTION_NUM_REQ, 0, NULL,
+    {"num-req", OPTION_NUM_REQ, 0, 0,
      "Num of requests to process, default -1 means all requests in the trace",
      1},
 
@@ -112,7 +112,7 @@ static struct argp_option options[] = {
 
     {NULL, 0, NULL, 0, "common parameters:", 0},
 
-    {"output", OPTION_OUTPUT_PATH, "", NULL, "Output path", 8},
+    {"output", OPTION_OUTPUT_PATH, "", OPTION_ARG_OPTIONAL, "Output path", 8},
     {"verbose", OPTION_VERBOSE, NULL, OPTION_ARG_OPTIONAL,
      "Produce verbose output", 8},
     {0}};

libCacheSim/bin/traceUtils/cli_parser.cpp

@@ -200,7 +200,7 @@ static void print_parsed_arg(struct arguments *args) {
   int n = 0;
   char output_str[OUTPUT_STR_LEN];
   n = snprintf(output_str, OUTPUT_STR_LEN - 1, "trace path: %s, trace_type %s",
-               args->trace_path, trace_type_str[args->trace_type]);
+               args->trace_path, g_trace_type_name[args->trace_type]);
 
   if (args->trace_type_params != NULL)
     n += snprintf(output_str + n, OUTPUT_STR_LEN - n - 1,

libCacheSim/bin/traceUtils/traceConv.cpp

@@ -96,7 +96,7 @@ void convert_to_oracleGeneral(reader_t *reader, std::string ofilepath,
           "%s: %ld M requests (%.2lf GB), trace time %ld, working set %lld "
           "object, %lld B (%.2lf GB)\n",
           reader->trace_path, (long)(n_req_curr / 1e6),
-          (double)total_bytes / GiB, start_ts - req->clock_time,
+          (double)total_bytes / GiB, (long)(start_ts - req->clock_time),
           (long long)n_obj, (long long)unique_bytes,
           (double)unique_bytes / GiB);
     }
@@ -118,8 +118,8 @@ void convert_to_oracleGeneral(reader_t *reader, std::string ofilepath,
       "%s: %ld M requests (%.2lf GB), trace time %ld, working set %lld "
       "object, %lld B (%.2lf GB), reversing output...\n",
       reader->trace_path, (long)(n_req_curr / 1e6), (double)total_bytes / GiB,
-      start_ts - req->clock_time, (long long)n_obj, (long long)unique_bytes,
-      (double)unique_bytes / GiB);
+      (long)(start_ts - req->clock_time), (long long)n_obj,
+      (long long)unique_bytes, (double)unique_bytes / GiB);
 
   struct trace_stat stat;
   stat.n_req = n_req_curr;
@@ -251,6 +251,6 @@ static void _reverse_file(std::string ofilepath, struct trace_stat stat,
   remove((ofilepath + ".reverse").c_str());
 
   INFO("trace conversion finished, %ld requests %ld objects, output %s\n",
-       n_req, stat.n_obj, ofilepath.c_str());
+       (long) n_req, (long) stat.n_obj, ofilepath.c_str());
 }
 }  // namespace traceConv

libCacheSim/bin/traceUtils/traceConvMain.cpp

@@ -30,7 +30,7 @@ int main(int argc, char *argv[]) {
   struct arguments args;
 
   cli::parse_cmd(argc, argv, &args);
-  if (args.ofilepath == NULL) {
+  if (strlen(args.ofilepath) == 0) {
     snprintf(args.ofilepath, OFILEPATH_LEN, "%s.oracleGeneral", args.trace_path);
   }
 

libCacheSim/bin/traceUtils/traceFilterMain.cpp

@@ -53,7 +53,7 @@ void filter(reader_t *reader, cache_t *cache, std::string ofilepath) {
     read_one_req(reader, req);
   }
 
-  INFO("write %ld/%ld %.4lf requests to file %s\n", n_written_req, n_req,
+  INFO("write %ld/%ld %.4lf requests to file %s\n", (long) n_written_req, (long) n_req,
        (double)n_written_req / n_req, ofilepath.c_str());
   free_request(req);
   output_file.close();

libCacheSim/cache/eviction/CR_LFU.c

@@ -46,7 +46,8 @@ static void free_list_node(void *list_node) {
  */
 cache_t *CR_LFU_init(const common_cache_params_t ccache_params,
                      const char *cache_specific_params) {
-  cache_t *cache = cache_struct_init("CR_LFU", ccache_params, cache_specific_params);
+  cache_t *cache =
+      cache_struct_init("CR_LFU", ccache_params, cache_specific_params);
   cache->cache_init = CR_LFU_init;
   cache->cache_free = CR_LFU_free;
   cache->get = CR_LFU_get;

libCacheSim/cache/eviction/LIRS.c

@@ -502,7 +502,7 @@ bool LIRS_can_insert(cache_t *cache, const request_t *req) {
 
   INFO("LIRS_can_insert: should not reach here. n_req = %ld\n",
        (long)cache->n_req);
-  assert(false);
+  abort();
 }
 
 static void LIRS_prune(cache_t *cache) {

libCacheSim/cache/eviction/WTinyLFU.c

@@ -253,6 +253,7 @@ cache_obj_t *WTinyLFU_insert(cache_t *cache, const request_t *req) {
 static cache_obj_t *WTinyLFU_to_evict(cache_t *cache, const request_t *req) {
   // Warning: don't use this function
   DEBUG_ASSERT(false);
+  return NULL;
 }
 
 static void WTinyLFU_evict(cache_t *cache, const request_t *req) {

libCacheSim/cache/eviction/nop.c

@@ -113,9 +113,9 @@ static bool nop_get(cache_t *cache, const request_t *req) {
  */
 static cache_obj_t *nop_find(cache_t *cache, const request_t *req,
                              const bool update_cache) {
-//   cache_obj_t *cache_obj = cache_find_base(cache, req, update_cache);
+  //   cache_obj_t *cache_obj = cache_find_base(cache, req, update_cache);
 
-//   return cache_obj;
+  //   return cache_obj;
   return NULL;
 }
 
@@ -130,9 +130,9 @@ static cache_obj_t *nop_find(cache_t *cache, const request_t *req,
  * @return the inserted object
  */
 static cache_obj_t *nop_insert(cache_t *cache, const request_t *req) {
-//   cache_obj_t *obj = cache_insert_base(cache, req);
+  //   cache_obj_t *obj = cache_insert_base(cache, req);
 
-//   return obj;
+  //   return obj;
   return NULL;
 }
 
@@ -147,7 +147,8 @@ static cache_obj_t *nop_insert(cache_t *cache, const request_t *req) {
  * @return the object to be evicted
  */
 static cache_obj_t *nop_to_evict(cache_t *cache, const request_t *req) {
-    assert(false);
+  assert(false);
+  return NULL;
 }
 
 /**
@@ -159,7 +160,7 @@ static cache_obj_t *nop_to_evict(cache_t *cache, const request_t *req) {
  * @param req not used
  */
 static void nop_evict(cache_t *cache, const request_t *req) {
-//   cache_evict_base(cache, obj_to_evict, true);
+  // cache_evict_base(cache, obj_to_evict, true);
 }
 
 /**

libCacheSim/cache/eviction/priv/QDLPv0.c

@@ -320,6 +320,7 @@ static cache_obj_t *QDLPv0_to_evict(cache_t *cache, const request_t *req) {
     // not implemented, we need to evict from the clock cache
     assert(0);
   }
+  return NULL;
 }
 
 /**