fix: allow MADV_HUGEPAGE/MADV_COLLAPSE and prlimit64 in gpu seccomp policy
NVIDIA's driver uses madvise with MADV_HUGEPAGE and MADV_COLLAPSE for GPU memory (scoped to render server only), and prlimit64 during initialization (added to gpu_common). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
parent
601d42aa36
commit
7ff2d31165
2 changed files with 4 additions and 0 deletions
|
|
@ -128,5 +128,6 @@ sched_get_priority_min: 1
|
|||
# Required for NVIDIA GPU
|
||||
mknodat: 1
|
||||
statfs: 1
|
||||
prlimit64: 1
|
||||
setsockopt: 1
|
||||
sendmmsg: 1
|
||||
|
|
|
|||
|
|
@ -16,3 +16,6 @@ socketpair: arg0 == AF_UNIX && arg1 == SOCK_SEQPACKET|SOCK_CLOEXEC && arg2 == 0
|
|||
|
||||
# allow signalfd()
|
||||
signalfd4: 1
|
||||
|
||||
# NVIDIA driver uses huge pages for GPU memory
|
||||
madvise: arg2 == MADV_HUGEPAGE || arg2 == MADV_COLLAPSE
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue