| 1 |
On Sat, Jan 06, 2018 at 11:26:43AM +1100, Adam Carter wrote |
| 2 |
> > |
| 3 |
> > > So, HAVE_EBPF_JIT=y just means that BPF JIT _can_ be done on x86. There |
| 4 |
> > > is a separate BPF_JIT setting to actually enable it. |
| 5 |
> > |
| 6 |
> > Well, that doesn't seem to be present here. Just the HAVE_ symbol. |
| 7 |
> |
| 8 |
> |
| 9 |
> Careful, there's BPF and EBPF. |
| 10 |
> |
| 11 |
> $ zgrep BPF /proc/config.gz |
| 12 |
> CONFIG_CGROUP_BPF=y |
| 13 |
> CONFIG_BPF=y |
| 14 |
> CONFIG_BPF_SYSCALL=y |
| 15 |
> # CONFIG_NETFILTER_XT_MATCH_BPF is not set |
| 16 |
> # CONFIG_NET_CLS_BPF is not set |
| 17 |
> # CONFIG_NET_ACT_BPF is not set |
| 18 |
> # CONFIG_BPF_JIT is not set |
| 19 |
> # CONFIG_BPF_STREAM_PARSER is not set |
| 20 |
> CONFIG_HAVE_EBPF_JIT=y |
| 21 |
> # CONFIG_TEST_BPF is not set |
| 22 |
|
| 23 |
I'm running openrc. On my 32-bit install, Intel Core2 duo, I get... |
| 24 |
|
| 25 |
zgrep BPF /proc/config.gz |
| 26 |
CONFIG_BPF=y |
| 27 |
# CONFIG_BPF_SYSCALL is not set |
| 28 |
# CONFIG_NETFILTER_XT_MATCH_BPF is not set |
| 29 |
# CONFIG_TEST_BPF is not set |
| 30 |
|
| 31 |
On my 64-bit install, Intel Silvermont (Atom), I get... |
| 32 |
|
| 33 |
zgrep BPF /proc/config.gz |
| 34 |
CONFIG_BPF=y |
| 35 |
# CONFIG_BPF_SYSCALL is not set |
| 36 |
# CONFIG_NETFILTER_XT_MATCH_BPF is not set |
| 37 |
# CONFIG_BPF_JIT is not set |
| 38 |
CONFIG_HAVE_EBPF_JIT=y |
| 39 |
# CONFIG_TEST_BPF is not set |
| 40 |
|
| 41 |
Does this improve security at all versus meltdown/spectre? Any |
| 42 |
suggestions for changes? |
| 43 |
|
| 44 |
-- |
| 45 |
Walter Dnes <waltdnes@××××××××.org> |
| 46 |
I don't run "desktop environments"; I run useful applications |
Archives