Repository navigation
comp: fix binding modules - #6773
Conversation
|
@fkwasowi can you check internal CI thanks ! |
There was a problem hiding this comment.
you'd also need to fix the error paths below. e_sink_connect label no longer needs to do an unbind right?
There was a problem hiding this comment.
Some whitespace error on L383 and L384, indentation seems off.
|
Is there an issue filed so we can see when this problem occurs? There is no issue created for this defect. The problem occurs, for example, during a test scenario \conv_fw_003_kd\test_01_kd_topologies |
kv2019i
left a comment
kv2019i
left a comment
There was a problem hiding this comment.
Looks like a good fix, but it would be helpful to explain why this has worked before and why the change is needed now. So maybe something like:
comp: fix module bind order
Current ipc_comp_connect() connects the source first, then calls comp_bind
and only then connects the sink. This leads to errors with modules that
assume sink to be connected at bind. This is a valid assumption, so fix
the order in ipc_comp_connect().
And some indentation issue in code. Maybe a github UI artifact, but please check.
There was a problem hiding this comment.
Some whitespace error on L383 and L384, indentation seems off.
c549d73 to
d375a53
Compare
There is no issue created for this defect. The problem occurs, for example, during a test scenario \conv_fw_003_kd\test_01_kd_topologies |
I have changed the description |
d375a53 to
85f4bf1
Compare
kv2019i
left a comment
kv2019i
left a comment
There was a problem hiding this comment.
Still some tuning in the error case, please see my inline comment.
There was a problem hiding this comment.
And have "e_source_bind" here.
There was a problem hiding this comment.
So actually, please keep this as "goto e_sink_connect;" and then
at end of function, have:
err:
comp_unbind(sink, bu);
e_sink_bind:
comp_unbind(source, bu);
e_src_bind:
pipeline_disconnect(sink, buffer, PPL_CONN_DIR_BUFFER_TO_COMP);
e_sink_connect:
pipeline_disconnect(source, buffer, PPL_CONN_DIR_COMP_TO_BUFFER);
There was a problem hiding this comment.
Done.Is it now correct?
85f4bf1 to
27fca6b
Compare
Current ipc_comp_connect() connects the source first, then calls comp_bind and only then connects the sink. This leads to errors with modules that assume sink to be connected at bind. This is a valid assumption, so fix the order in ipc_comp_connect(). Signed-off-by: Kwasowiec, Fabiola <fabiola.kwasowiec@intel.com>
27fca6b to
6da6515
Compare
|
SOFCI TEST |
|
Patch looks good to me. I have tested on WoV pipeline. This can fix #6784 |
Current ipc_comp_connect() connects the source first, then calls comp_bind
and only then connects the sink. This leads to errors with modules that
assume sink to be connected at bind. This is a valid assumption, so fix
the order in ipc_comp_connect().