Attention is currently required from: jolly, laforge.
3 comments:
Commit Message:
Patch Set #3, Line 17: This change ensures that newly received data is appended to the existing
Maybe I should state that I don't want to reuse the pending buffer for the read operation. […]
This is still confusing to me, because you are still using the term "pending buffer" twice.
IIUC on the second one you mean "to the buffer where next read will be done by the kernel"?
File src/core/osmo_io.c:
Patch Set #3, Line 178: struct msgb *iofd_msgb_alloc2(struct osmo_io_fd *iofd, size_t size)
The application should only care about the length of data in the buffer not the size.
I disagree. Apps may want to make sure a msgb allocated is of at least a certain size, so that they can reuse the msgb with zero-copy by appending/prepending to it, etc.
If you say you garantee the msgb recevied by the user is of at least osmo_iofd msgb_size configured by the user, then I'm fine, otherwise we need to revisit this.
Patch Set #3, Line 335: * If the pending message is not large enough, create a larger message. */
The application should only care about the length of data in the buffer not the size.
My understanding is that since user expects no segment bigger than the configured msgb_size, then we should read() only up to the maximum msgb_size configured. Hence it can never happen that we read more than the size of the msgb.
To view, visit change 40584. To unsubscribe, or for help writing mail filters, visit settings.