1 files changed, 169 insertions, 56 deletions

diff --git a/lib/zstd/compress/zstd_cwksp.h b/lib/zstd/compress/zstd_cwksp.h
index 98e359adf5d4..349fc923c355 100644
--- a/lib/zstd/compress/zstd_cwksp.h
+++ b/lib/zstd/compress/zstd_cwksp.h
@@ -32,6 +32,10 @@
 #define ZSTD_CWKSP_ASAN_REDZONE_SIZE 128
 #endif
 
+
+/* Set our tables and aligneds to align by 64 bytes */
+#define ZSTD_CWKSP_ALIGNMENT_BYTES 64
+
 /*-*************************************
 *  Structures
 ***************************************/
@@ -114,10 +118,11 @@ typedef enum {
  * - Tables: these are any of several different datastructures (hash tables,
  *   chain tables, binary trees) that all respect a common format: they are
  *   uint32_t arrays, all of whose values are between 0 and (nextSrc - base).
- *   Their sizes depend on the cparams.
+ *   Their sizes depend on the cparams. These tables are 64-byte aligned.
  *
  * - Aligned: these buffers are used for various purposes that require 4 byte
- *   alignment, but don't require any initialization before they're used.
+ *   alignment, but don't require any initialization before they're used. These
+ *   buffers are each aligned to 64 bytes.
  *
  * - Buffers: these buffers are used for various purposes that don't require
  *   any alignment or initialization before they're used. This means they can
@@ -130,8 +135,7 @@ typedef enum {
  *
  * 1. Objects
  * 2. Buffers
- * 3. Aligned
- * 4. Tables
+ * 3. Aligned/Tables
  *
  * Attempts to reserve objects of different types out of order will fail.
  */
@@ -184,6 +188,8 @@ MEM_STATIC size_t ZSTD_cwksp_align(size_t size, size_t const align) {
  * Since tables aren't currently redzoned, you don't need to call through this
  * to figure out how much space you need for the matchState tables. Everything
  * else is though.
+ *
+ * Do not use for sizing aligned buffers. Instead, use ZSTD_cwksp_aligned_alloc_size().
  */
 MEM_STATIC size_t ZSTD_cwksp_alloc_size(size_t size) {
     if (size == 0)
@@ -191,66 +197,139 @@ MEM_STATIC size_t ZSTD_cwksp_alloc_size(size_t size) {
     return size;
 }
 
-MEM_STATIC void ZSTD_cwksp_internal_advance_phase(
-        ZSTD_cwksp* ws, ZSTD_cwksp_alloc_phase_e phase) {
+/*
+ * Returns an adjusted alloc size that is the nearest larger multiple of 64 bytes.
+ * Used to determine the number of bytes required for a given "aligned".
+ */
+MEM_STATIC size_t ZSTD_cwksp_aligned_alloc_size(size_t size) {
+    return ZSTD_cwksp_alloc_size(ZSTD_cwksp_align(size, ZSTD_CWKSP_ALIGNMENT_BYTES));
+}
+
+/*
+ * Returns the amount of additional space the cwksp must allocate
+ * for internal purposes (currently only alignment).
+ */
+MEM_STATIC size_t ZSTD_cwksp_slack_space_required(void) {
+    /* For alignment, the wksp will always allocate an additional n_1=[1, 64] bytes
+     * to align the beginning of tables section, as well as another n_2=[0, 63] bytes
+     * to align the beginning of the aligned section.
+     *
+     * n_1 + n_2 == 64 bytes if the cwksp is freshly allocated, due to tables and
+     * aligneds being sized in multiples of 64 bytes.
+     */
+    size_t const slackSpace = ZSTD_CWKSP_ALIGNMENT_BYTES;
+    return slackSpace;
+}
+
+
+/*
+ * Return the number of additional bytes required to align a pointer to the given number of bytes.
+ * alignBytes must be a power of two.
+ */
+MEM_STATIC size_t ZSTD_cwksp_bytes_to_align_ptr(void* ptr, const size_t alignBytes) {
+    size_t const alignBytesMask = alignBytes - 1;
+    size_t const bytes = (alignBytes - ((size_t)ptr & (alignBytesMask))) & alignBytesMask;
+    assert((alignBytes & alignBytesMask) == 0);
+    assert(bytes != ZSTD_CWKSP_ALIGNMENT_BYTES);
+    return bytes;
+}
+
+/*
+ * Internal function. Do not use directly.
+ * Reserves the given number of bytes within the aligned/buffer segment of the wksp,
+ * which counts from the end of the wksp (as opposed to the object/table segment).
+ *
+ * Returns a pointer to the beginning of that space.
+ */
+MEM_STATIC void*
+ZSTD_cwksp_reserve_internal_buffer_space(ZSTD_cwksp* ws, size_t const bytes)
+{
+    void* const alloc = (BYTE*)ws->allocStart - bytes;
+    void* const bottom = ws->tableEnd;
+    DEBUGLOG(5, "cwksp: reserving %p %zd bytes, %zd bytes remaining",
+        alloc, bytes, ZSTD_cwksp_available_space(ws) - bytes);
+    ZSTD_cwksp_assert_internal_consistency(ws);
+    assert(alloc >= bottom);
+    if (alloc < bottom) {
+        DEBUGLOG(4, "cwksp: alloc failed!");
+        ws->allocFailed = 1;
+        return NULL;
+    }
+    /* the area is reserved from the end of wksp.
+     * If it overlaps with tableValidEnd, it voids guarantees on values' range */
+    if (alloc < ws->tableValidEnd) {
+        ws->tableValidEnd = alloc;
+    }
+    ws->allocStart = alloc;
+    return alloc;
+}
+
+/*
+ * Moves the cwksp to the next phase, and does any necessary allocations.
+ * cwksp initialization must necessarily go through each phase in order.
+ * Returns a 0 on success, or zstd error
+ */
+MEM_STATIC size_t
+ZSTD_cwksp_internal_advance_phase(ZSTD_cwksp* ws, ZSTD_cwksp_alloc_phase_e phase)
+{
     assert(phase >= ws->phase);
     if (phase > ws->phase) {
+        /* Going from allocating objects to allocating buffers */
         if (ws->phase < ZSTD_cwksp_alloc_buffers &&
                 phase >= ZSTD_cwksp_alloc_buffers) {
             ws->tableValidEnd = ws->objectEnd;
         }
+
+        /* Going from allocating buffers to allocating aligneds/tables */
         if (ws->phase < ZSTD_cwksp_alloc_aligned &&
                 phase >= ZSTD_cwksp_alloc_aligned) {
-            /* If unaligned allocations down from a too-large top have left us
-             * unaligned, we need to realign our alloc ptr. Technically, this
-             * can consume space that is unaccounted for in the neededSpace
-             * calculation. However, I believe this can only happen when the
-             * workspace is too large, and specifically when it is too large
-             * by a larger margin than the space that will be consumed. */
-            /* TODO: cleaner, compiler warning friendly way to do this??? */
-            ws->allocStart = (BYTE*)ws->allocStart - ((size_t)ws->allocStart & (sizeof(U32)-1));
-            if (ws->allocStart < ws->tableValidEnd) {
-                ws->tableValidEnd = ws->allocStart;
+            {   /* Align the start of the "aligned" to 64 bytes. Use [1, 64] bytes. */
+                size_t const bytesToAlign =
+                    ZSTD_CWKSP_ALIGNMENT_BYTES - ZSTD_cwksp_bytes_to_align_ptr(ws->allocStart, ZSTD_CWKSP_ALIGNMENT_BYTES);
+                DEBUGLOG(5, "reserving aligned alignment addtl space: %zu", bytesToAlign);
+                ZSTD_STATIC_ASSERT((ZSTD_CWKSP_ALIGNMENT_BYTES & (ZSTD_CWKSP_ALIGNMENT_BYTES - 1)) == 0); /* power of 2 */
+                RETURN_ERROR_IF(!ZSTD_cwksp_reserve_internal_buffer_space(ws, bytesToAlign),
+                                memory_allocation, "aligned phase - alignment initial allocation failed!");
             }
-        }
+            {   /* Align the start of the tables to 64 bytes. Use [0, 63] bytes */
+                void* const alloc = ws->objectEnd;
+                size_t const bytesToAlign = ZSTD_cwksp_bytes_to_align_ptr(alloc, ZSTD_CWKSP_ALIGNMENT_BYTES);
+                void* const objectEnd = (BYTE*)alloc + bytesToAlign;
+                DEBUGLOG(5, "reserving table alignment addtl space: %zu", bytesToAlign);
+                RETURN_ERROR_IF(objectEnd > ws->workspaceEnd, memory_allocation,
+                                "table phase - alignment initial allocation failed!");
+                ws->objectEnd = objectEnd;
+                ws->tableEnd = objectEnd;  /* table area starts being empty */
+                if (ws->tableValidEnd < ws->tableEnd) {
+                    ws->tableValidEnd = ws->tableEnd;
+        }   }   }
         ws->phase = phase;
+        ZSTD_cwksp_assert_internal_consistency(ws);
     }
+    return 0;
 }
 
 /*
  * Returns whether this object/buffer/etc was allocated in this workspace.
  */
-MEM_STATIC int ZSTD_cwksp_owns_buffer(const ZSTD_cwksp* ws, const void* ptr) {
+MEM_STATIC int ZSTD_cwksp_owns_buffer(const ZSTD_cwksp* ws, const void* ptr)
+{
     return (ptr != NULL) && (ws->workspace <= ptr) && (ptr <= ws->workspaceEnd);
 }
 
 /*
  * Internal function. Do not use directly.
  */
-MEM_STATIC void* ZSTD_cwksp_reserve_internal(
-        ZSTD_cwksp* ws, size_t bytes, ZSTD_cwksp_alloc_phase_e phase) {
+MEM_STATIC void*
+ZSTD_cwksp_reserve_internal(ZSTD_cwksp* ws, size_t bytes, ZSTD_cwksp_alloc_phase_e phase)
+{
     void* alloc;
-    void* bottom = ws->tableEnd;
-    ZSTD_cwksp_internal_advance_phase(ws, phase);
-    alloc = (BYTE *)ws->allocStart - bytes;
-
-    if (bytes == 0)
+    if (ZSTD_isError(ZSTD_cwksp_internal_advance_phase(ws, phase)) || bytes == 0) {
         return NULL;
+    }
 
 
-    DEBUGLOG(5, "cwksp: reserving %p %zd bytes, %zd bytes remaining",
-        alloc, bytes, ZSTD_cwksp_available_space(ws) - bytes);
-    ZSTD_cwksp_assert_internal_consistency(ws);
-    assert(alloc >= bottom);
-    if (alloc < bottom) {
-        DEBUGLOG(4, "cwksp: alloc failed!");
-        ws->allocFailed = 1;
-        return NULL;
-    }
-    if (alloc < ws->tableValidEnd) {
-        ws->tableValidEnd = alloc;
-    }
-    ws->allocStart = alloc;
+    alloc = ZSTD_cwksp_reserve_internal_buffer_space(ws, bytes);
 
 
     return alloc;
@@ -259,33 +338,44 @@ MEM_STATIC void* ZSTD_cwksp_reserve_internal(
 /*
  * Reserves and returns unaligned memory.
  */
-MEM_STATIC BYTE* ZSTD_cwksp_reserve_buffer(ZSTD_cwksp* ws, size_t bytes) {
+MEM_STATIC BYTE* ZSTD_cwksp_reserve_buffer(ZSTD_cwksp* ws, size_t bytes)
+{
     return (BYTE*)ZSTD_cwksp_reserve_internal(ws, bytes, ZSTD_cwksp_alloc_buffers);
 }
 
 /*
- * Reserves and returns memory sized on and aligned on sizeof(unsigned).
+ * Reserves and returns memory sized on and aligned on ZSTD_CWKSP_ALIGNMENT_BYTES (64 bytes).
  */
-MEM_STATIC void* ZSTD_cwksp_reserve_aligned(ZSTD_cwksp* ws, size_t bytes) {
-    assert((bytes & (sizeof(U32)-1)) == 0);
-    return ZSTD_cwksp_reserve_internal(ws, ZSTD_cwksp_align(bytes, sizeof(U32)), ZSTD_cwksp_alloc_aligned);
+MEM_STATIC void* ZSTD_cwksp_reserve_aligned(ZSTD_cwksp* ws, size_t bytes)
+{
+    void* ptr = ZSTD_cwksp_reserve_internal(ws, ZSTD_cwksp_align(bytes, ZSTD_CWKSP_ALIGNMENT_BYTES),
+                                            ZSTD_cwksp_alloc_aligned);
+    assert(((size_t)ptr & (ZSTD_CWKSP_ALIGNMENT_BYTES-1))== 0);
+    return ptr;
 }
 
 /*
- * Aligned on sizeof(unsigned). These buffers have the special property that
+ * Aligned on 64 bytes. These buffers have the special property that
  * their values remain constrained, allowing us to re-use them without
  * memset()-ing them.
  */
-MEM_STATIC void* ZSTD_cwksp_reserve_table(ZSTD_cwksp* ws, size_t bytes) {
+MEM_STATIC void* ZSTD_cwksp_reserve_table(ZSTD_cwksp* ws, size_t bytes)
+{
     const ZSTD_cwksp_alloc_phase_e phase = ZSTD_cwksp_alloc_aligned;
-    void* alloc = ws->tableEnd;
-    void* end = (BYTE *)alloc + bytes;
-    void* top = ws->allocStart;
+    void* alloc;
+    void* end;
+    void* top;
+
+    if (ZSTD_isError(ZSTD_cwksp_internal_advance_phase(ws, phase))) {
+        return NULL;
+    }
+    alloc = ws->tableEnd;
+    end = (BYTE *)alloc + bytes;
+    top = ws->allocStart;
 
     DEBUGLOG(5, "cwksp: reserving %p table %zd bytes, %zd bytes remaining",
         alloc, bytes, ZSTD_cwksp_available_space(ws) - bytes);
     assert((bytes & (sizeof(U32)-1)) == 0);
-    ZSTD_cwksp_internal_advance_phase(ws, phase);
     ZSTD_cwksp_assert_internal_consistency(ws);
     assert(end <= top);
     if (end > top) {
@@ -296,27 +386,31 @@ MEM_STATIC void* ZSTD_cwksp_reserve_table(ZSTD_cwksp* ws, size_t bytes) {
     ws->tableEnd = end;
 
 
+    assert((bytes & (ZSTD_CWKSP_ALIGNMENT_BYTES-1)) == 0);
+    assert(((size_t)alloc & (ZSTD_CWKSP_ALIGNMENT_BYTES-1))== 0);
     return alloc;
 }
 
 /*
  * Aligned on sizeof(void*).
+ * Note : should happen only once, at workspace first initialization
  */
-MEM_STATIC void* ZSTD_cwksp_reserve_object(ZSTD_cwksp* ws, size_t bytes) {
-    size_t roundedBytes = ZSTD_cwksp_align(bytes, sizeof(void*));
+MEM_STATIC void* ZSTD_cwksp_reserve_object(ZSTD_cwksp* ws, size_t bytes)
+{
+    size_t const roundedBytes = ZSTD_cwksp_align(bytes, sizeof(void*));
     void* alloc = ws->objectEnd;
     void* end = (BYTE*)alloc + roundedBytes;
 
 
-    DEBUGLOG(5,
+    DEBUGLOG(4,
         "cwksp: reserving %p object %zd bytes (rounded to %zd), %zd bytes remaining",
         alloc, bytes, roundedBytes, ZSTD_cwksp_available_space(ws) - roundedBytes);
-    assert(((size_t)alloc & (sizeof(void*)-1)) == 0);
-    assert((bytes & (sizeof(void*)-1)) == 0);
+    assert((size_t)alloc % ZSTD_ALIGNOF(void*) == 0);
+    assert(bytes % ZSTD_ALIGNOF(void*) == 0);
     ZSTD_cwksp_assert_internal_consistency(ws);
     /* we must be in the first phase, no advance is possible */
     if (ws->phase != ZSTD_cwksp_alloc_objects || end > ws->workspaceEnd) {
-        DEBUGLOG(4, "cwksp: object alloc failed!");
+        DEBUGLOG(3, "cwksp: object alloc failed!");
         ws->allocFailed = 1;
         return NULL;
     }
@@ -328,7 +422,8 @@ MEM_STATIC void* ZSTD_cwksp_reserve_object(ZSTD_cwksp* ws, size_t bytes) {
     return alloc;
 }
 
-MEM_STATIC void ZSTD_cwksp_mark_tables_dirty(ZSTD_cwksp* ws) {
+MEM_STATIC void ZSTD_cwksp_mark_tables_dirty(ZSTD_cwksp* ws)
+{
     DEBUGLOG(4, "cwksp: ZSTD_cwksp_mark_tables_dirty");
 
 
@@ -451,6 +546,24 @@ MEM_STATIC int ZSTD_cwksp_reserve_failed(const ZSTD_cwksp* ws) {
 *  Functions Checking Free Space
 ***************************************/
 
+/* ZSTD_alignmentSpaceWithinBounds() :
+ * Returns if the estimated space needed for a wksp is within an acceptable limit of the
+ * actual amount of space used.
+ */
+MEM_STATIC int ZSTD_cwksp_estimated_space_within_bounds(const ZSTD_cwksp* const ws,
+                                                        size_t const estimatedSpace, int resizedWorkspace) {
+    if (resizedWorkspace) {
+        /* Resized/newly allocated wksp should have exact bounds */
+        return ZSTD_cwksp_used(ws) == estimatedSpace;
+    } else {
+        /* Due to alignment, when reusing a workspace, we can actually consume 63 fewer or more bytes
+         * than estimatedSpace. See the comments in zstd_cwksp.h for details.
+         */
+        return (ZSTD_cwksp_used(ws) >= estimatedSpace - 63) && (ZSTD_cwksp_used(ws) <= estimatedSpace + 63);
+    }
+}
+
+
 MEM_STATIC size_t ZSTD_cwksp_available_space(ZSTD_cwksp* ws) {
     return (size_t)((BYTE*)ws->allocStart - (BYTE*)ws->tableEnd);
 }