aboutsummaryrefslogtreecommitdiffstats
path: root/rules.mk
diff options
context:
space:
mode:
authorFelix Fietkau <nbd@nbd.name>2018-07-12 17:19:07 +0200
committerFelix Fietkau <nbd@nbd.name>2018-07-13 16:39:49 +0200
commit16a9ddfe64e2dc80328c3b7ac81d5d87a7da9233 (patch)
treea736b9332c93db2d0b9292c6f5ec33522d8c3651 /rules.mk
parent0e6cfb6919149267202c6dbaf9e5aab3ab6d6b1d (diff)
downloadupstream-16a9ddfe64e2dc80328c3b7ac81d5d87a7da9233.tar.gz
upstream-16a9ddfe64e2dc80328c3b7ac81d5d87a7da9233.tar.bz2
upstream-16a9ddfe64e2dc80328c3b7ac81d5d87a7da9233.zip
ramips: ethernet: use own page_frag_cache
Using the NAPI or netdev frag cache along with other drivers can lead to 32 KiB pages being held for a long time, despite only being used for very few page fragment. This can happen if the ethernet driver grabs one or two fragments for rx ring refill, while other drivers use (and free up) the remaining fragments. The 32 KiB higher-order page can only be freed once all users have freed their fragments, which only happens after the rings of all drivers holding the fragments have wrapped around. Depending on the traffic patterns, this can waste a lot of memory and look a lot like a memory leak Signed-off-by: Felix Fietkau <nbd@nbd.name>
Diffstat (limited to 'rules.mk')
0 files changed, 0 insertions, 0 deletions