vulkan packing code clean, drop image storage type

[nihui]

No description provided.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 87.55187% with 30 lines in your changes missing coverage. Please review.

Project coverage is 95.81%. Comparing base (bd0b111) to head (6146309).
Report is 1 commits behind head on master.

Files with missing lines	Patch %	Lines
src/layer/vulkan/packing_vulkan.cpp	85.64%	30 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #6128      +/-   ##
==========================================
- Coverage   95.81%   95.81%   -0.01%     
==========================================
  Files         831      831              
  Lines      264935   264990      +55     
==========================================
+ Hits       253843   253895      +52     
- Misses      11092    11095       +3     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[github-actions]

The binary size change of libncnn.so (bytes)

architecture	base size	pr size	difference
x86_64	15656872	15598184	-58688 😘
armhf	6665712	6607084	-58628 😘
aarch64	9985232	9917304	-67928 😘

[Copilot]

Pull Request Overview

Cleanup Vulkan packing code by removing image‐storage variants, unifying on a new cast_type model, and updating layer and GPU APIs to drop storage_type.

• Removed legacy *_fp*_to_fp* shader files and storage‐type branches in shaders
• Renamed storage_type_from/to to cast_type_from/to in Packing and shader code
• Updated convert_packing overloads and get_utility_operator to work with cast_type only

Reviewed Changes

Copilot reviewed 34 out of 34 changed files in this pull request and generated 3 comments.

File	Description
src/layer/vulkan/shader/packing_pack4to1.comp	Unified on cast_type, removed storage‐type branching
src/layer/packing.h	Dropped storage_type fields, updated comments for cast_type
src/gpu.cpp	Added convert_packing overload, refactored get_utility_operator
src/command.cpp	Updated calls to convert_packing to pass cast_type_to

Comments suppressed due to low confidence (1)

src/gpu.cpp:4226

• The new convert_packing overload with cast_type_to introduces multiple cast branches (auto, fp32, fp16). Add unit or integration tests to cover each cast_type_to path to ensure correct behavior.

void VulkanDevice::convert_packing(const VkMat& src, VkMat& dst, int dst_elempack, int cast_type_to, VkCompute& cmd, const Option& opt) const