Hi,
With some larger mesh networks running Gluon some Out-of-Memory cases on 32MB RAM devices were reported [0].
For the following Mem-Info section, the number of free pages confuses me a little:
[...] [ 236.529242] Mem-Info: [ 236.531561] active_anon:691 inactive_anon:9 isolated_anon:0 [ 236.531561] active_file:131 inactive_file:138 isolated_file:0 [ 236.531561] unevictable:0 dirty:0 writeback:0 unstable:0 [ 236.531561] slab_reclaimable:205 slab_unreclaimable:3105 [ 236.531561] mapped:39 shmem:28 pagetables:80 bounce:0 [ 236.531561] free:289 free_pcp:0 free_cma:0 [ 236.563259] Normal free:1156kB min:1024kB low:1280kB high:1536kB active_anon:2764kB inactive_anon:36kB active_file:524kB inactive_file:552kB unevictable:0kB isolated(anon):0kB isolated(file):0kB present:32768kB managed:27776kB mlocked:0kB dirty:0kB writeback:0kB mapped:156kB shmem:112kB slab_reclaimable:820kB slab_unreclaimable:12420kB kernel_stack:424kB pagetables:320kB unstable:0kB bounce:0kB free_pcp:0kB local_pcp:0kB free_cma:0kB writeback_tmp:0kB pages_scanned:6492 all_unreclaimable? yes [ 236.607070] lowmem_reserve[]: 0 0 [ 236.610425] Normal: 125*4kB (UME) 34*8kB (UME) 16*16kB (UME) 4*32kB (ME) 0*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 1156kB [ 236.623157] 297 total pagecache pages [ 236.626852] 0 pages in swap cache [ 236.630189] Swap cache stats: add 0, delete 0, find 0/0 [ 236.635455] Free swap = 0kB [ 236.638355] Total swap = 0kB [ 236.641251] 8192 pages RAM [ 236.643971] 0 pages HighMem/MovableOnly [ 236.647837] 1248 pages reserved [...]
(Full log here[1], including some slabinfo; with a backport of these slabinfo related patches applied [2])
It says that only 289 pages are left. However those numbers listed in the Mem-Info section do not seem to add up to 8192 minus 1289 reserved minus 289 free pages.
Any idea why it might list only 289 free pages?
Regards, Linus
[0]: https://github.com/freifunk-gluon/gluon/issues/1243 https://github.com/freifunk-gluon/gluon/issues/1197 [1]: https://paste.linuxlounge.net/#/GUMP_cQOKwgqQhf5_gBfPU1Vv0A!7bT80KaCMPcUC_zI... [2]: https://lkml.org/lkml/2017/10/4/788