diff --git a/src/broadcom/compiler/nir_to_vir.c b/src/broadcom/compiler/nir_to_vir.c
index 3a2465fbe10..01d920cb470 100644
--- a/src/broadcom/compiler/nir_to_vir.c
+++ b/src/broadcom/compiler/nir_to_vir.c
@@ -1901,7 +1901,14 @@ v3d_optimize_nir(struct v3d_compile *c, struct nir_shader *s)
                         .callback = mem_vectorize_callback,
                         .robust_modes = 0,
                 };
-                NIR_PASS(progress, s, nir_opt_load_store_vectorize, &vectorize_opts);
+                bool vectorize_progress = false;
+                NIR_PASS(vectorize_progress, s, nir_opt_load_store_vectorize,
+                         &vectorize_opts);
+                if (vectorize_progress) {
+                        NIR_PASS(progress, s, nir_lower_alu_to_scalar, NULL, NULL);
+                        NIR_PASS(progress, s, nir_lower_pack);
+                        progress = true;
+                }
 
                 if (lower_flrp != 0) {
                         bool lower_flrp_progress = false;
diff --git a/src/broadcom/vulkan/v3dv_pipeline.c b/src/broadcom/vulkan/v3dv_pipeline.c
index 2a7ba0ee34b..10694cc180d 100644
--- a/src/broadcom/vulkan/v3dv_pipeline.c
+++ b/src/broadcom/vulkan/v3dv_pipeline.c
@@ -215,6 +215,9 @@ const nir_shader_compiler_options v3dv_nir_options = {
    .lower_unpack_snorm_4x8 = true,
    .lower_pack_half_2x16 = true,
    .lower_unpack_half_2x16 = true,
+   .lower_pack_32_2x16 = true,
+   .lower_pack_32_2x16_split = true,
+   .lower_unpack_32_2x16_split = true,
    /* FIXME: see if we can avoid the uadd_carry and usub_borrow lowering and
     * get the tests to pass since it might produce slightly better code.
     */
diff --git a/src/gallium/drivers/v3d/v3d_screen.c b/src/gallium/drivers/v3d/v3d_screen.c
index 1864b95be0e..3a9019c070f 100644
--- a/src/gallium/drivers/v3d/v3d_screen.c
+++ b/src/gallium/drivers/v3d/v3d_screen.c
@@ -705,6 +705,9 @@ static const nir_shader_compiler_options v3d_nir_options = {
         .lower_unpack_snorm_4x8 = true,
         .lower_pack_half_2x16 = true,
         .lower_unpack_half_2x16 = true,
+        .lower_pack_32_2x16 = true,
+        .lower_pack_32_2x16_split = true,
+        .lower_unpack_32_2x16_split = true,
         .lower_fdiv = true,
         .lower_find_lsb = true,
         .lower_ffma16 = true,