From b3f594771f49ba0db6012a03c7a29e87f3eaec0f Mon Sep 17 00:00:00 2001
From: Patrick Vogler <patrick.vogler@hlrs.de>
Date: Fri, 28 Jun 2024 11:27:58 +0200
Subject: [PATCH] Previously the library accepted a max of 255 threads for
 compression. The limit has now been extended to 9 223 372 036 854 775 807,
 which should suffice for the next couple of years.

---
 src/library/dwt.c   | 16 ++++++++++++----
 src/library/tier1.c | 19 +++++++++++--------
 2 files changed, 23 insertions(+), 12 deletions(-)

diff --git a/src/library/dwt.c b/src/library/dwt.c
index 025439a..b22c7f2 100755
--- a/src/library/dwt.c
+++ b/src/library/dwt.c
@@ -1624,18 +1624,22 @@ forward_wavelet_transform(bwc_field *const field, bwc_parameter *const parameter
    uint64   rX1, rY1, rZ1;
    uint64   width, height, depth;
    uint64   x, y, z;
-   uint32   buff_size;
+
+   int64    nThreads;
    int16    i;
+   
+   uint32   buff_size;
+   
    uint16   incr_TS;
    uint16   rTS0;
    uint16   rTS1;
    uint16   dt;
    uint16   t;
+   
    uint8    id;
    uint8    filter_tapsX, filter_tapsY, filter_tapsZ;
    uint8    filter_tapsTS;
    uint8    level;
-   uint8    nThreads;
 
    /*-----------------------*\
    ! DEFINE STRUCTS:         !
@@ -2161,18 +2165,22 @@ inverse_wavelet_transform(bwc_field *const field, bwc_parameter *const parameter
    uint64   rX1, rY1, rZ1;
    uint64   width, height, depth;
    uint64   x, y, z;
+   
+   int64    nThreads;
+   int64    i;
+   
    uint32   buff_size;
-   int16    i;
+   
    uint16   incr_TS;
    uint16   rTS0;
    uint16   rTS1;
    uint16   dt;
    uint16   t;
+   
    uint8    id;
    uint8    filter_tapsX, filter_tapsY, filter_tapsZ;
    uint8    filter_tapsTS;
    uint8    level;
-   uint8    nThreads;
 
    /*-----------------------*\
    ! DEFINE STRUCTS:         !
diff --git a/src/library/tier1.c b/src/library/tier1.c
index ff160b7..60abda2 100755
--- a/src/library/tier1.c
+++ b/src/library/tier1.c
@@ -2836,12 +2836,14 @@ t1_encode(bwc_field *const field, bwc_tile *const tile, bwc_parameter *const par
    uint64   c;
    uint64   cbSizeX, cbSizeY, cbSizeZ;
    uint64   width, height, depth;
+
    int64    buff_size;
-   int64    j;
+   int64    i, j;
+   int64    nThreads;
+
    uint16   cbSizeTS;
    uint16   slope_max, slope_min;
-   int16    i, z;
-   uint8    nThreads;
+   int16    z;
 
    /*-----------------------*\
    ! DEFINE STRUCTS:         !
@@ -2921,7 +2923,7 @@ t1_encode(bwc_field *const field, bwc_tile *const tile, bwc_parameter *const par
    ! threads during a parallel run. For a serial run only one !
    ! working buffer is allocated.                             !
    \*--------------------------------------------------------*/
-   for(i = 0; i < nThreads; ++i)
+   for(i = 0; i < (int64)nThreads; ++i)
    {
       memory[i] = calloc(buff_size, sizeof(bwc_coder_stripe));
       if(!memory[i])
@@ -3137,11 +3139,12 @@ t1_decode(bwc_field *const field, bwc_tile *const tile, bwc_parameter *const par
    uint64   c;
    uint64   cbSizeX, cbSizeY, cbSizeZ;
    uint64   width, height, depth;
+
    int64    buff_size;
-   int64    j;
+   int64    i, j;
+   int64    nThreads;
+
    uint16   cbSizeTS;
-   int16    i;
-   uint8    nThreads;
 
    /*-----------------------*\
    ! DEFINE STRUCTS:         !
@@ -3214,7 +3217,7 @@ t1_decode(bwc_field *const field, bwc_tile *const tile, bwc_parameter *const par
    ! threads during a parallel run. For a serial run only one !
    ! working buffer is allocated.                             !
    \*--------------------------------------------------------*/
-   for(i = 0; i < nThreads; ++i)
+   for(i = 0; i < (int64)nThreads; ++i)
    {
       memory[i] = calloc(buff_size, sizeof(bwc_coder_stripe));
       if(!memory[i])