radv: fix UB in radv_format_pack_clear_color for snorm formats

Casting a negative float to uint64_t is undefined behavior. GCC 15 with
-O2 produces 0xFFFFFFFFFFFFFFFF for (uint64_t)(-32767.5f), causing snorm
clear values to be packed incorrectly (e.g. 0xFFFF instead of 0x8001 for
snorm16 -1.0). This results in wrong DCC comp-to-single clear colors and
~966 CTS snorm multisample_resolve test failures.

Fix by casting through int64_t first, which is well-defined (truncation
toward zero) and preserves the two's complement bit pattern.

Fixes: 585c25be1e ("radv: fix color conversions for normalized uint/sint formats")
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
(cherry picked from commit 2595940b0d)

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41269>

.pick_status.json

@@ -1114,7 +1114,7 @@
         "description": "radv: fix UB in radv_format_pack_clear_color for snorm formats",
         "nominated": false,
         "nomination_type": 2,
-        "resolution": 4,
+        "resolution": 1,
         "main_sha": null,
         "because_sha": "585c25be1e1a96568829542f2765e97c63d79939",
         "notes": null

src/amd/vulkan/radv_formats.c

@@ -592,7 +592,7 @@ radv_format_pack_clear_color(VkFormat format, uint32_t clear_vals[2], VkClearCol
             else
                f -= 0.5f;
 
-            v = (uint64_t)f;
+            v = (uint64_t)(int64_t)f;
          }
       } else if (channel->type == UTIL_FORMAT_TYPE_FLOAT) {
          if (channel->size == 32) {