* add int8 definitions
* packing vulkan int8/int32, quantize vulkan
* vulkan dequantize
* requantize vulkan

* fix pyncnn build

* always prefer xtheadvector
* update ci toolchain

vkimagemat was originally used as a mat storage in the hope of improving performance on old adreno gpus, but in fact it is slower than the cpu in most cases and is no longer suitable for the latest adreno architecture and large shapes

* fix sub mat cstep
* fix embed
* rnn/lstm/gru int8 test without rounding diversity

* fix innerproduct gemm vulkan

Co-authored-by: mmyyy22 <mmyyy22@users.noreply.github.com>
Co-authored-by: nihui <nihuini@tencent.com>

Replaces `setenv` with `putenv` because `setenv` is not available on
some toolchains (e.g., llvm-mingw, clang-cl), leading to compilation
errors.

* warn puttext
* split dict

* fix memory leak
* fuse conv+bias early

* pnnx update batch 1 mask attention, add more clip test

* fuse same qkv input

* skip onnx clip sdpa

* maintain cache

* local cache

* skip transformer test on torch<2.1

* zip pnnx

* bool8 to bool

* initialize xlnet weights

* onnx einsum

* onnx reducelogsumexp

* skip tensor as scalar attribute

* skip longformer