yzh119 commented on code in PR #14161:
URL: https://github.com/apache/tvm/pull/14161#discussion_r1122696620
##########
include/tvm/tir/schedule/schedule.h:
##########
@@ -405,6 +405,36 @@ class ScheduleNode : public runtime::Object {
virtual BlockRV CacheWrite(const BlockRV& block_rv, int write_buffer_index,
const String& storage_scope,
const Array<BlockRV> consumer_blocks = {}) = 0;
+ /*!
+ * \brief Create a block that reads a buffer region into a read cache. It
requires:
+ * 1) There is at most one block who writes the buffer in the scope.
+ * 2) The scope block have stage-pipeline property.
+ * Compared to cache read, the indices to access allocated cache buffer is
customized by user.
+ * \param block_rv The consumer block of the target buffer.
+ * \param read_buffer_index The index of the buffer in block's read region.
+ * \param storage_scope The target storage scope.
+ * \param index_map User defined indices to access allocated cache buffer,
maps from block iter
+ * vars. \param consumer_blocks An optional list of consumers to read from
cache directly. \return
+ * The cache stage block.
+ */
Review Comment:
Hi, my original code looks like what you suggested but it doesn't pass CI,
see
[here](https://ci.tlcpack.ai/blue/organizations/jenkins/tvm-lint/detail/PR-14161/4/pipeline),
and by default, clang-format would format the code like this:
```
* \param index_map User defined indices to access allocated cache buffer,
maps from block iter
* vars. \param consumer_blocks An optional list of consumers to read from
cache directly. \return
```
do you have any suggestions on how to change the default behavior of
clang-format and CI?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
