vulkan: use scalar FA rather than coopmat2 when N==1

[jeffbolznv]

before:
Z:\github\jeffbolznv\llama.cpp\build\bin\RelWithDebInfo>llama-bench.exe -m C:\models\llama-2-7b.Q4_0.gguf -n 128,1024,4096,8192 -p 0 -fa 0,1 --prio 2  --repetitions 1
ggml_vulkan: Found 1 Vulkan devices:
ggml_vulkan: 0 = NVIDIA GeForce RTX 4070 (NVIDIA) | uma: 0 | fp16: 1 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2
| model                          |       size |     params | backend    | ngl | fa |            test |                  t/s |
| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | --------------: | -------------------: |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  0 |           tg128 |        100.81 ± 0.00 |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  0 |          tg1024 |         94.63 ± 0.00 |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  0 |          tg4096 |         77.88 ± 0.00 |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  0 |          tg8192 |         62.73 ± 0.00 |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  1 |           tg128 |         98.30 ± 0.00 |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  1 |          tg1024 |         93.27 ± 0.00 |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  1 |          tg4096 |         77.11 ± 0.00 |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  1 |          tg8192 |         62.65 ± 0.00 |

after:
Z:\github\jeffbolznv\llama.cpp\build\bin\RelWithDebInfo>llama-bench.exe -m C:\models\llama-2-7b.Q4_0.gguf -n 128,1024,4096,8192 -p 0 -fa 0,1 --prio 2  --repetitions 1
ggml_vulkan: Found 1 Vulkan devices:
ggml_vulkan: 0 = NVIDIA GeForce RTX 4070 (NVIDIA) | uma: 0 | fp16: 1 | warp size: 32 | shared memory: 49152 | int dot: 1 | matrix cores: NV_coopmat2
| model                          |       size |     params | backend    | ngl | fa |            test |                  t/s |
| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | --------------: | -------------------: |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  0 |           tg128 |        100.92 ± 0.00 |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  0 |          tg1024 |         94.62 ± 0.00 |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  0 |          tg4096 |         77.38 ± 0.00 |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  0 |          tg8192 |         62.65 ± 0.00 |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  1 |           tg128 |         98.89 ± 0.00 |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  1 |          tg1024 |         94.21 ± 0.00 |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  1 |          tg4096 |         80.34 ± 0.00 |
| llama 7B Q4_0                  |   3.56 GiB |     6.74 B | Vulkan     |  99 |  1 |          tg8192 |         66.53 ± 0.00 |