large vector sizes failure - cpu compilation - quantised models

[PhaneeshB]

What happened?

On compiling a model with int8 quantization one of the dispatches fails to compile with the following error:

error: One or more operations with large vector sizes (16384 bytes) were found 

Min repro adapted from the failing dispatch:

module {
  func.func @largeVectorMinRepro(%arg0: tensor<1x320x65x65xi8>) -> tensor<1x320x1x1xf32> {
        %cst = arith.constant 1.250000e-01 : f32
        %cst_0 = arith.constant 0.000000e+00 : f32
        %c5408000 = arith.constant 5408000 : index
        %c0 = arith.constant 0 : index
        %3 = tensor.empty() : tensor<1x320x1x1xf32>
        %4 = tensor.empty() : tensor<65x65xf32>
        %5 = tensor.empty() : tensor<1x320x65x65xf32>
        %6 = linalg.fill ins(%cst_0 : f32) outs(%3 : tensor<1x320x1x1xf32>) -> tensor<1x320x1x1xf32>
        %7 = linalg.generic {indexing_maps = [affine_map<(d0, d1, d2, d3) -> (d0, d1, d2, d3)>, affine_map<(d0, d1, d2, d3) -> (d0, d1, d2, d3)>], iterator_types = ["parallel", "parallel", "parallel", "parallel"]} ins(%arg0 : tensor<1x320x65x65xi8>) outs(%5 : tensor<1x320x65x65xf32>) {
        ^bb0(%in: i8, %out: f32):
          %9 = arith.extsi %in : i8 to i32
          %10 = arith.sitofp %9 : i32 to f32
          %11 = arith.mulf %10, %cst : f32
          linalg.yield %11 : f32
        } -> tensor<1x320x65x65xf32>
        %8 = linalg.pooling_nchw_sum  ins(%7, %4 : tensor<1x320x65x65xf32>, tensor<65x65xf32>) outs(%6 : tensor<1x320x1x1xf32>) -> tensor<1x320x1x1xf32>
    return %8 : tensor<1x320x1x1xf32>
  }
}

compile command : iree-compile --iree-input-demote-i64-to-i32 --iree-hal-target-backends=llvm-cpu largevectorissue.minrepro.mlir -o test.vmfb

host issue here

Steps to reproduce your issue

1. Go to '...'
2. Click on '....'
3. Scroll down to '....'
4. See error

What component(s) does this issue relate to?

No response

Version information

No response

Additional context

No response