Document changes to benchmarking and print number of threads for disp…

…lay level >= 4

programs/zstd.1.md

@@ -343,7 +343,7 @@ If the value of `ZSTD_CLEVEL` is not a valid integer, it will be ignored with a
 
 `ZSTD_NBTHREADS` can be used to set the number of threads `zstd` will attempt to use during compression.
 If the value of `ZSTD_NBTHREADS` is not a valid unsigned integer, it will be ignored with a warning message.
-`ZSTD_NBTHREADS` has a default value of (`min(4, nbCores/4`), and is capped at ZSTDMT_NBWORKERS_MAX==200.
+`ZSTD_NBTHREADS` has a default value of (`min(4, nbCores/4)`), and is capped at ZSTDMT_NBWORKERS_MAX==200.
 `zstd` must be compiled with multithread support for this variable to have any effect.
 
 They can both be overridden by corresponding command line arguments:
@@ -664,9 +664,12 @@ BENCHMARK
 The `zstd` CLI provides a benchmarking mode that can be used to easily find suitable compression parameters, or alternatively to benchmark a computer's performance.
 `zstd -b [FILE(s)]` will benchmark `zstd` for both compression and decompression using default compression level.
 Note that results are very dependent on the content being compressed.
+
 It's possible to pass multiple files to the benchmark, and even a directory with `-r DIRECTORY`.
 When no `FILE` is provided, the benchmark will use a procedurally generated `lorem ipsum` text.
 
+Benchmarking will employ `min(4, nbCores/4)` worker threads by default in order to match the behavior of the normal CLI I/O.
+
 * `-b#`:
     benchmark file(s) using compression level #
 * `-e#`:

programs/zstdcli.c

@@ -21,8 +21,7 @@
 #endif
 
 #ifndef ZSTDCLI_NBTHREADS_DEFAULT
-#define ZSTDCLI_NBTHREADS_DEFAULT                                              \
-  UTIL_countLogicalCores() / 4 > 4 ? 4 : UTIL_countLogicalCores() / 4
+#define ZSTDCLI_NBTHREADS_DEFAULT (UTIL_countLogicalCores() / 4 > 4 ? 4 : UTIL_countLogicalCores() / 4)
 #endif
 
 /*-************************************
@@ -761,7 +760,7 @@ static int init_cLevel(void) {
 }
 
 #ifdef ZSTD_MULTITHREAD
-static unsigned init_nbThreads(void) {
+static unsigned default_nbThreads(void) {
     const char* const env = getenv(ENV_NBTHREADS);
     if (env != NULL) {
         const char* ptr = env;
@@ -856,7 +855,7 @@ int main(int argCount, const char* argv[])
     ZSTD_paramSwitch_e mmapDict=ZSTD_ps_auto;
     ZSTD_paramSwitch_e useRowMatchFinder = ZSTD_ps_auto;
     FIO_compressionType_t cType = FIO_zstdCompression;
-    unsigned nbWorkers = 0;
+    int nbWorkers = -1; /* -1 means unset */
     double compressibility = -1.0;  /* lorem ipsum generator */
     unsigned bench_nbSeconds = 3;   /* would be better if this value was synchronized from bench */
     size_t blockSize = 0;
@@ -897,17 +896,13 @@ int main(int argCount, const char* argv[])
 #endif
     ZSTD_paramSwitch_e literalCompressionMode = ZSTD_ps_auto;
 
-
     /* init */
     checkLibVersion();
     (void)recursive; (void)cLevelLast;    /* not used when ZSTD_NOBENCH set */
     (void)memLimit;
     assert(argCount >= 1);
     if ((filenames==NULL) || (file_of_names==NULL)) { DISPLAYLEVEL(1, "zstd: allocation error \n"); exit(1); }
     programName = lastNameFromPath(programName);
-#ifdef ZSTD_MULTITHREAD
-    nbWorkers = init_nbThreads();
-#endif
 
     /* preset behaviors */
     if (exeNameMatch(programName, ZSTD_ZSTDMT)) nbWorkers=0, singleThread=0;
@@ -1299,7 +1294,7 @@ int main(int argCount, const char* argv[])
     DISPLAYLEVEL(3, WELCOME_MESSAGE);
 
 #ifdef ZSTD_MULTITHREAD
-    if ((operation==zom_decompress) && (!singleThread) && (nbWorkers > 1)) {
+    if ((operation==zom_decompress) && (nbWorkers > 1)) {
         DISPLAYLEVEL(2, "Warning : decompression does not support multi-threading\n");
     }
     if ((nbWorkers==0) && (!singleThread)) {
@@ -1312,6 +1307,15 @@ int main(int argCount, const char* argv[])
             DISPLAYLEVEL(3, "Note: %d physical core(s) detected \n", nbWorkers);
         }
     }
+    // Resolve to default if nbWorkers is still unset
+    if (nbWorkers == -1) {
+      if (operation == zom_decompress) {
+        nbWorkers = 1;
+      } else {
+        nbWorkers = default_nbThreads();
+      }
+    }
+    DISPLAYLEVEL(4, "Compressing with %u worker threads \n", nbWorkers);
 #else
     (void)singleThread; (void)nbWorkers; (void)defaultLogicalCores;
 #endif