wip: new: runtime tracepoints management#577
wip: new: runtime tracepoints management#577FedeDP wants to merge 1 commit intocleanup/unify_simpleconsumer_simpledriverfrom
Conversation
poiana
added
release-note-none
kind/feature
FedeDP
changed the base branch from
master
to
cleanup/unify_simpleconsumer_simpledriver
|
This PR was branched from #521. |
| } | ||
| else | ||
| { | ||
| if (ret == 0) |
There was a problem hiding this comment.
This is a small bug fix on top of #521 that is not actually needed in that PR, as tracepoints are only set at startup time and we don't need g_tracepoints_attached to be actually synced (and nobody will ever touch 'g_tracepoints_attached` at runtime).
| if(put_user(PPM_SCHEMA_CURRENT_VERSION, out)) | ||
| ret = -EINVAL; | ||
| goto cleanup_ioctl_nolock; | ||
| } else if (cmd == PPM_IOCTL_GET_TPMASK) { |
There was a problem hiding this comment.
New ioctl to get current g_tracepoints_attached.
We could also cache this info userspace side, but i think driver is always more accurate (ie: if some tracepoints could not be attached/detached, that is reflected on the kmod side g_tracepoints_attached but not in the userspace-side requested tp_of_interest).
| size_t m_ncpus; | ||
| char* m_lasterr; | ||
| int m_bpf_prog_fds[BPF_PROGS_MAX]; | ||
| struct bpf_prog m_bpf_progs[BPF_PROGS_MAX]; |
There was a problem hiding this comment.
Merged prog_fds and event_fd; plus, store tracepoint name for further usage.
| int m_bpf_event_fd[BPF_PROGS_MAX]; | ||
| int m_bpf_map_fds[BPF_MAPS_MAX]; | ||
| int m_bpf_prog_array_map_idx; | ||
| char m_filepath[PATH_MAX]; |
There was a problem hiding this comment.
Store probe filepath; used in scap_bpf.
| { | ||
| if (strcmp(handle->m_bpf_progs[i].name, shname) == 0) | ||
| { | ||
| already_attached = true; |
There was a problem hiding this comment.
Check if the requested tracepoint is already attached (ie: if we already have its name in m_bpf_progs).
Avoid attaching it twice.
| ASSERT(false); | ||
| return SCAP_FAILURE; | ||
| uint32_t op = oargs->tp_of_interest.tp[i] ? SCAP_TPMASK_SET : SCAP_TPMASK_UNSET; | ||
| scap_kmod_handle_tp_mask(engine, op, i); |
There was a problem hiding this comment.
Use the new function here.
| } | ||
|
|
||
| switch(tp) | ||
| static int32_t update_single_tp_of_interest(int tp, bool interesting) |
There was a problem hiding this comment.
New API to manage a single tp of interest.
| break; | ||
| } | ||
|
|
||
| if (ppm_sc >= PPM_SC_MAX) |
There was a problem hiding this comment.
Moved the checks at scap level: there were indeed no checks here, and one could provide invalid data.
| break; | ||
| } | ||
|
|
||
| if (tp >= TP_VAL_MAX) |
| { | ||
| throw sinsp_exception("you cannot use this method before opening the inspector!"); | ||
| } | ||
| if (ppm_sc >= PPM_SC_MAX || ppm_sc < 0) |
There was a problem hiding this comment.
Managed at scap level now.
This is the same API we introduced for runtime ppm_sc management in #521, but for tracepoints. It is implemented for kmod, bpf and modern_bpf. Signed-off-by: Federico Di Pierro <nierro92@gmail.com>
FedeDP
force-pushed
the
new/runtime_tp_management
branch
from
4f94c74 to
5054494
Compare
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: FedeDP The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
/reopen |
|
@FedeDP: Failed to re-open PR: state cannot be changed. The cleanup/unify_simpleconsumer_simpledriver branch has been deleted. DetailsIn response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
2 participants
What type of PR is this?
/kind feature
Any specific area of the project related to this PR?
/area driver-kmod
/area driver-bpf
/area driver-modern-bpf
/area libscap-engine-bpf
/area libscap-engine-kmod
/area libscap-engine-modern-bpf
/area libscap
/area libsinsp
Does this PR require a change in the driver versions?
/version driver-API-version-major
new ioctl introduced in kmod.
What this PR does / why we need it:
This PR completes part of #521 allowing even tracepoints to be runtime modified, just like we allow it for ppm_sc.
Which issue(s) this PR fixes:
Fixes #
Special notes for your reviewer:
Does this PR introduce a user-facing change?: