Reworking IO Ring pipeline to fully support multiple infilght files

Reworking the filequeue, the buffer chaining logic and the error
handling.
Renaming functions.
Fix bug in arena.

.gitignore

@@ -9,3 +9,4 @@ temp_code.c
 /.cache/clangd/index
 /file_hasher
 /io_uring_test
+/file_hasher

arena.c

@@ -437,12 +437,14 @@ void *arena_push(mem_arena **arena_ptr, u64 size, bool zero) { // mk push
      Commit memory if needed
      ------------------------------------------------------------ */
 
-  if (local_post > selected->commit_pos) {
+  if (local_post > selected->commit_pos -
-    u64 new_commit = ALIGN_UP_POW2(local_post, arena_pagesize());
+                       ALIGN_UP_POW2(sizeof(mem_arena), selected->align)) {
+    u64 new_commit = ALIGN_UP_POW2(local_post + ALIGN_UP_POW2(sizeof(mem_arena), selected->align), arena_pagesize());
     new_commit = MIN(new_commit, selected->reserve_size);
 
     if (!plat_mem_commit((u8 *)selected + selected->commit_pos,
                          new_commit - selected->commit_pos)) {
+        printf("ERROR: Could not commit memory!\n");
       return NULL;
     }
 

base.h

@@ -46,6 +46,7 @@
 #include <stdlib.h>
 #include <string.h>
 #include <time.h>
+#include <limits.h>
 
 /* ------------------------------------------------------------
    Base types

binaries/changelog.txt

@@ -50,7 +50,7 @@ Fixing user prompt parsing
 Reorganising the code
 Improving the scan function
 
-5.0: Implementing the IO Ring for windows and ui_uring for linux instead of buffered hashing, huge performance gains. The IO Ring is event driven, thread local, uses DMA and direct disk I/O, bypassing the OS cash completely, registred buffers, it supports bashing multiple submissions and can handle multiple files at the same time.
+5.0: Implementing the IO Ring for windows and ui_uring for linux instead of buffered hashing, huge performance gains. The IO Ring is event driven, thread local, uses DMA and direct disk I/O, bypassing the OS cache completely, registred buffers, it supports bashing multiple submissions and can handle multiple files at the same time.
 Hashing small files using XXH3_128bits() instead of the streaming pipeline(XXH3_128bits_reset(), XXH3_128bits_update(), XXH3_128bits_digest()), this reduses the overhead of creating a state and digest, coupled with the IO Ring it improves the hashing of small files whose size is inferior to the size of IO Ring buffers
 fixing the xxh_x86dispatch warnings
 Updating the progress printing function

file_hasher.c

@@ -77,14 +77,19 @@ int main(int argc, char **argv) {
   // Detect hardware
   // -------------------------------
   // --- Windows: detect PHYSICAL cores (not logical threads) ---
-  size_t hw_threads = platform_physical_cores();
+  uint32_t cpu_cores = platform_physical_cores();
 
   // Logical threads = CPU cores * 2
-  size_t num_threads = hw_threads * 2;
+  uint32_t cpu_threads = cpu_cores * 2;
 
-  printf("Starting thread pool: %zu threads (CPU cores: %zu)\n", num_threads,
+  uint32_t num_scan_threads = cpu_threads;
-         hw_threads);
+  uint32_t num_hash_threads = cpu_threads;
-  printf("  Selected instruction set: %s\n", get_xxhash_instruction_set());
+  // uint32_t num_hash_threads = 1;
+
+  printf("%d cores %d threads CPU detected with %s instruction set\n"
+         "Starting thread pool: %d scanning and %d hashing threads\n",
+         cpu_cores, cpu_threads, get_xxhash_instruction_set(), num_scan_threads,
+         num_hash_threads);
 
   // Align IO Ring block size to the system page size
   g_ioring_buffer_size = ALIGN_UP_POW2(g_ioring_buffer_size, g_pagesize);
@@ -119,9 +124,6 @@ int main(int argc, char **argv) {
   // }
 
   // Starting hash threads
-  size_t num_hash_threads = num_threads;
-  // size_t num_hash_threads = 1;
-
   WorkerContext workers[num_hash_threads];
   Thread *hash_threads =
       arena_push(&gp_arena, sizeof(Thread) * num_hash_threads, true);
@@ -130,7 +132,7 @@ int main(int argc, char **argv) {
     workers[i].arena = arena_create(&params);
     workers[i].file_queue = &file_queue;
 
-    if (thread_create(&hash_threads[i], (ThreadFunc)hash_worker_io_ring,
+    if (thread_create(&hash_threads[i], (ThreadFunc)hash_worker_ioring,
                       &workers[i]) != 0) {
       fprintf(stderr, "Failed to create hash thread %zu\n", i);
       exit(1);
@@ -178,8 +180,6 @@ int main(int argc, char **argv) {
   }
 
   // Starting scan threads
-  size_t num_scan_threads = num_threads;
-
   ScannerContext scanners[num_scan_threads];
   Thread *scan_threads =
       arena_push(&gp_arena, sizeof(Thread) * num_scan_threads, true);