use size_t type for proper large tensor support

[nihui]

No description provided.

[tencent-adm]

Thank you for your submission, we really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[Copilot]

Pull Request Overview

This PR modernizes tensor dimension handling by changing key variables from int to size_t to properly support large tensors. This is important for avoiding integer overflow when dealing with large matrices and tensors where dimensions can exceed the range of 32-bit integers.

Key changes include:

• Converting stride variables from int to size_t in softmax implementations
• Updating tensor step calculations to use size_t for memory offsets
• Changing matrix dimension variables in GEMM operations to size_t

Reviewed Changes

Copilot reviewed 33 out of 33 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
src/layer/x86/softmax_x86.cpp	Updates stride parameters to size_t in softmax functions
src/layer/x86/gemm_x86.cpp	Changes tensor step calculations to size_t for GEMM operations
src/layer/x86/gemm_int8.h	Updates matrix step variables to size_t in int8 GEMM functions
src/layer/x86/convolution_packed_int8.h	Changes tensor step calculations to size_t in convolution
src/layer/x86/convolution_packed.h	Updates tensor step variables to size_t
src/layer/x86/convolution_im2col_gemm_int8.h	Changes offset calculations to size_t
src/layer/x86/convolution_3x3_winograd_int8.h	Updates tensor step variables to size_t
src/layer/x86/convolution_3x3_winograd.h	Changes tensor step calculations to size_t
src/layer/slice.cpp	Updates size calculations to size_t
src/layer/riscv/softmax_riscv.cpp	Changes stride parameters to size_t
src/layer/riscv/gemm_riscv.cpp	Updates tensor step variables to size_t
src/layer/glu.cpp	Changes offset calculation to size_t
src/layer/gemm.cpp	Updates tensor step calculations to size_t
src/layer/arm/*	Multiple ARM-specific files with similar size_t conversions

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 95.70%. Comparing base (fe509e9) to head (2a9cf5c).

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #6201      +/-   ##
==========================================
- Coverage   96.17%   95.70%   -0.47%     
==========================================
  Files         835      835              
  Lines      260827   265247    +4420     
==========================================
+ Hits       250843   253860    +3017     
- Misses       9984    11387    +1403     

☔ 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-advanced-security]

CodeQL found more than 20 potential problems in the proposed changes. Check the Files changed tab for more details.

[github-actions]

The binary size change of libncnn.so (bytes)

architecture	base size	pr size	difference
x86_64	15639200	15643128	+3928 ⚠️
armhf	6648076	6648220	+144 ⚠️
aarch64	9987568	9986896	-672 😘