We implemented the feature to fill unassembled slots w/dummy header
values for the livox pcloud header.
We also fixed a bug where io uring was writing into the last slot
only because we were using the same iovec for every SQE.
We call stop() inside the assembleFrameReq3, so when it returns,
the eventfdDesc should be destroyed. Don't allow a possibly stale
eventfdDesc obj to permit us to re-arm the eventfdDesc read_some
call.
This is in preparation for re-using StagingBuffer to also serve
as the collation buffer that we'll use as the intermediate stage
for producing the final output mesh.
Invoke it instimFrameProductionTimesliceInd.
Also, we discovered:
* stream_descriptor::release() doesn't fully release all metadata
from the fd it was assigned. This suggests that we should go
through the codebase and do: release()=>reset() whenever we
wish to release().
* We've confirmed that spinlocks can be used to prevent race
conditions between stop() and handler methods.
StagingBuffer:
We now allocate memory with mmap(MAP_ANONYMOUS) so that we can be
sure it can be pinned with io_uring_register_buffers(). This
ensures that if DMA is possible, it should be usable.
IoUringAssemblyEngine:
We now register an eventfd with io_uring so that we can listen
for CQEs with boost::asio.
This symbol is defined as a static member object inside of a
boost detail header. When boost headers are used in a project
that uses Boost in both the main binary as well as dlopen()'d
shlibs, the top_ symbol gets duplicated and the metadata gets
partitioned.
We use the Boost shlib to unify both the main binary and the
shlibs to use the same memory address for top_.
This involves marking the templated object call_stack::top_ as
"extern" and then declaring to Boost that we intend to use the
shlibs.
We move the methods in StimulusBuffer whose addresses are taken during
program execution into a separate static lib. This guarantees that
they'll have their own, single vaddr at runtime, at least within
each independent code module.
hayodea
latentprion