The netpoll subsystem pre-allocates 32 SKBs in a pool for emergency use
during out-of-memory conditions. However, the current implementation has
several inefficiencies:
* The SKB pool, once allocated, is never freed:
* Resources remain allocated even after netpoll users are removed
* Failed initialization can leave pool populated forever
* The global pool design makes resource tracking difficult
This series addresses these issues through three patches:
Patch 1 ("net: netpoll: Individualize the skb pool"):
- Replace global pool with per-user pools in netpoll struct
Patch 2 ("net: netpoll: flush skb pool during cleanup"):
- Properly free pool resources during netconsole cleanup
These changes improve resource management and make the code more
maintainable. As a side benefit, the improved structure would allow
netpoll to be modularized if desired in the future.
What is coming next?
Once this patch is integrated, I am planning to have the SKBs being
refilled outside of hot (send) path, in a work thread.
Signed-off-by: Breno Leitao <leitao@debian.org>
---
Changelog:
v3:
* Use the skb_queue_purge_reason() helper instead of dropping the skb
pool manually. (Jakub)
v2:
* Drop the very first patch from v1 ("net: netpoll: Defer skb_pool
population until setup success") (Jakub)
* Move skb_queue_head_init() to the first patch, where it belongs to
(Jakub)
* https://lore.kernel.org/all/20241107-skb_buffers_v2-v2-0-288c6264ba4f@debian.org/
v1:
* https://lore.kernel.org/all/20241025142025.3558051-1-leitao@debian.org/
- Link to v2: https://lore.kernel.org/r/20241107-skb_buffers_v2-v2-0-288c6264ba4f@debian.org
---
Breno Leitao (2):
net: netpoll: Individualize the skb pool
net: netpoll: flush skb pool during cleanup
include/linux/netpoll.h | 1 +
net/core/netpoll.c | 45 ++++++++++++++++++++++++++-------------------
2 files changed, 27 insertions(+), 19 deletions(-)
---
base-commit: a71c69f51d1119db5f7812b35f16e8ef7786b3f2
change-id: 20241107-skb_buffers_v2-f3e626100eda
Best regards,
--
Breno Leitao <leitao@debian.org>