Difference between revisions of "Long lists"
Jump to navigation
Jump to search
(Created page with "Collected here is the list of, well, lists :) that might be too long to be iterated over in a plain manner * parasite_collect_aios looks for AIO ring VMA which is typically t...") |
(minor formatting) |
||
| Line 1: | Line 1: | ||
| − | Collected here is the list of, well, lists :) that might be too long to be iterated over in a plain manner | + | Collected here is the list of, well, lists :) that might be too long to be iterated over in a plain manner. |
* parasite_collect_aios looks for AIO ring VMA which is typically the only one in MM | * parasite_collect_aios looks for AIO ring VMA which is typically the only one in MM | ||
| Line 15: | Line 15: | ||
== See also == | == See also == | ||
| − | [[Performance research]] | + | * [[Performance research]] |
| + | |||
[[Category: Development]] | [[Category: Development]] | ||
Latest revision as of 21:31, 13 February 2017
Collected here is the list of, well, lists :) that might be too long to be iterated over in a plain manner.
- parasite_collect_aios looks for AIO ring VMA which is typically the only one in MM
- find_fle_by_fd iterates over fdtable FDs to look for a specific one
- create_children_and_session makes 2 scans, while we can do it in one go
- opts.external will grow big soon (external_lookup_*)
- collect_one_fifo scans fifo_head for object by ID
- restore_file_locks scans global list for per-pid locks
- find_file_desc_raw uses 64-chains hash. Make it dynamic?
- rsti->used seems to be non-efficient
- opts.inherit_fds will grow big soon
- find_inotify_info & collect_fanotify_mark
- inet_ports in port_add
- find_unix_sk_by_ino