[WIP] add maximum limit for grid of reduce, elementwise and masked_select kernel

[FlyingQianMM]

PR types

Bug fixes

PR changes

APIs

Describe

The grid number of reduce、elementwise and masked_select has not been limited, which may raise a bug like:

:parallel_for failed: cudaErrorInvalidConfiguration: invalid configuration argument. 

So we add a maximum limit for grid of reduce, elementwise and masked_select kernel.

该pr已经拆成三个pr合入：
1、add maximum limit for grid of reduce, elementwise, gather and scatter： https://github.com/PaddlePaddle/Paddle/pull/40813/files
2、support get_item where the index is a bool scalar tensor： #40829
3、support gather when index is 0-zero shape: #40952

[paddle-bot-old]

Thanks for your contribution!
Please wait for the result of CI firstly. See Paddle CI Manual for details.

[paddle-bot-old]

Sorry to inform you that 7d91e3c's CIs have passed for more than 7 days. To prevent PR conflicts, you need to re-run all CIs manually.

[paddle-bot-old]

Sorry to inform you that 6073788's CIs have passed for more than 7 days. To prevent PR conflicts, you need to re-run all CIs manually.