1 |
On Tue, 30 Oct 2018 13:29:59 +0100 Håkon Alstadheim wrote: |
2 |
> |
3 |
> Den 30. okt. 2018 10:01, skrev Mick: |
4 |
> > On Tuesday, 30 October 2018 06:30:23 GMT Håkon Alstadheim wrote: |
5 |
> >> I'm having fun enabling "test" in FEATURES on my gentoo-desktop. One |
6 |
> >> interesting failure, that brings to mind build failures I have had in |
7 |
> >> the past: |
8 |
> >> |
9 |
> >> Building sys-apps/mlocate-0.26-r2, I get |
10 |
> >> |
11 |
> >> 43: updatedb: Very deep hierarchy FAILED |
12 |
> >> (updatedb.at:261) |
13 |
> >> |
14 |
> >> Trying to reproduce, as root I do "make check" in the work/mlocate-0.26/ |
15 |
> >> , and the test passes. |
16 |
> >> |
17 |
> >> 43: updatedb: Very deep hierarchy ok |
18 |
> >> |
19 |
> >> I'd really like to get to the bottom of this, as I believe it must have |
20 |
> >> the same root-cause as issues I have had compiling large packages such |
21 |
> >> as firefox. |
22 |
> >> |
23 |
> >> Re-running both the emerge and the make check, I get the same results. |
24 |
> >> emerge fails, make check succeeds. I made a local copy of the ebuild and |
25 |
> >> inserted a "ulimit -a" in pre_src_test. |
26 |
> >> |
27 |
> >> ulimit from root-shell: |
28 |
> >> |
29 |
> >> # ulimit -a |
30 |
> >> core file size (blocks, -c) unlimited |
31 |
> >> data seg size (kbytes, -d) unlimited |
32 |
> >> scheduling priority (-e) 0 |
33 |
> >> file size (blocks, -f) unlimited |
34 |
> >> pending signals (-i) 59958 |
35 |
> >> max locked memory (kbytes, -l) 16384 |
36 |
> >> max memory size (kbytes, -m) unlimited |
37 |
> >> open files (-n) 1024 |
38 |
> >> pipe size (512 bytes, -p) 8 |
39 |
> >> POSIX message queues (bytes, -q) 819200 |
40 |
> >> real-time priority (-r) 0 |
41 |
> >> stack size (kbytes, -s) 8192 |
42 |
> >> cpu time (seconds, -t) unlimited |
43 |
> >> max user processes (-u) 10000 |
44 |
> >> virtual memory (kbytes, -v) unlimited |
45 |
> >> file locks (-x) unlimited |
46 |
> >> |
47 |
> >> ulimit from emerge: |
48 |
> >>>>> Source compiled. |
49 |
> >> core file size (blocks, -c) unlimited |
50 |
> >> data seg size (kbytes, -d) unlimited |
51 |
> >> scheduling priority (-e) 0 |
52 |
> >> file size (blocks, -f) unlimited |
53 |
> >> pending signals (-i) 59958 |
54 |
> >> max locked memory (kbytes, -l) 16384 |
55 |
> >> max memory size (kbytes, -m) unlimited |
56 |
> >> open files (-n) 1024 |
57 |
> >> pipe size (512 bytes, -p) 8 |
58 |
> >> POSIX message queues (bytes, -q) 819200 |
59 |
> >> real-time priority (-r) 0 |
60 |
> >> stack size (kbytes, -s) 9788 |
61 |
> >> cpu time (seconds, -t) unlimited |
62 |
> >> max user processes (-u) 10000 |
63 |
> >> virtual memory (kbytes, -v) unlimited |
64 |
> >> file locks (-x) unlimited |
65 |
> >> |
66 |
> >>>>> Test phase: sys-apps/mlocate-0.26-r2 |
67 |
> >> I have plenty of space in my portage temp directory (/pt): |
68 |
> >> |
69 |
> >> # df -hT ./ |
70 |
> >> Filsystem Type Størrelse Brukt Tilgj. Bruk% Montert på |
71 |
> >> /dev/xvdc ext4 163G 8,0G 147G 6% /pt |
72 |
> >> |
73 |
> >> Portage temp is at /pt due to the earlier mentioned issues with firefox. |
74 |
> >> |
75 |
> >> At my wits end here. Anyone ? |
76 |
> > I have not looked or used the test FEATURES of portage, but have also noticed |
77 |
> > over time certain packages fail to build on machines with low RAM. As low |
78 |
> > here I consider anything less than 4G. It seems each thread is now consuming |
79 |
> > so much memory they cumulatively use up all RAM available and then start |
80 |
> > swapping endlessly until the compilation invariably fails. Increasingly more |
81 |
> > and more packages have been suffering from this, the last two I noticed are |
82 |
> > qtwebkit and qtwebengine. |
83 |
> > |
84 |
> > My solution has been to create a package.env file in which I specify MAKEOPTS |
85 |
> > limiting the number of jobs and average load for any of these packages which |
86 |
> > chew up all the RAM. |
87 |
> Memory should not be a problem here. Fails with only that one emerge |
88 |
> running, |
89 |
> succeeds if run directly as root, or with FEATURES="-sandbox -usersandbox". |
90 |
> |
91 |
> Memory is >14GB: |
92 |
> # vmstat |
93 |
> procs -----------memory---------- ---swap-- -----io---- -system-- |
94 |
> ------cpu----- |
95 |
> r b swpd free buff cache si so bi bo in cs us sy |
96 |
> id wa st |
97 |
> 3 4 28416 6904608 174112 4616144 0 0 65 266 13 4 10 |
98 |
> 2 84 4 0 |
99 |
|
100 |
It is possible that you hit directory loop. What lstree says on |
101 |
that dir? Anyway, report this to sandbox devs. |
102 |
|
103 |
Best regards, |
104 |
Andrew Savchenko |