We don't need a socketpair for this. Additionally, things seem to work
better from the service when the client provides the pipe. Otherwise, when
running as a dbus service I often have issues with things getting closed
out from under us.
This allows you to specify --stack-size=(multiple_of_page_size) to unwind
from captured stack contents. It will use the new SysprofUserSampler to
unwind stack traces via sysprof-live-unwinder.
This instrument triggers the live unwinder in sysprofd to capture a
pre-configured amount of stack contents and CPU registers. You can use
this instead of SysprofSampler in cases where you do not have frame-
pointers but want a useful trace.
It does have a moderate amount of CPU overhead compared to just relying
on frame-pointers so keep that in mind. Generally useful on platforms
that do not have frame pointers such as CentOS.
This provides a new sysprof-live-unwinder subprocess that runs as root to
allow accessing all processes on the system via /proc/$pid/. It is spawned
by sysprofd with various perf event FDs and a FD to write captures to.
Ideally the capture_fd is something that will naturally error if the client
application crashes (such as a socketpair() having the peer close). This
is not enforced but encouraged. Additionally, an event_fd is used to allow
the client application to signal the live-unwinder to exit.
Unwinding is performed by looking at the modules loaded into the target
pid and using libdwfl to access DWARF/CFI/etc state machinery. Stack data
does not touch the disk as it exists in a mmap buffer from perf and is
then translated into a callchain and sent to the Sysprof client.
Unwinding occurs as normal post-mortem though is improved through the use
of debuginfod to locate the appropriate symbols.
This requires a coordinating sysprofd that knows how to handle reading the
new attributes. Setting these fields will allow snapshotting the contents
of the stack and registers to do offline unwinding.
Also make the conversion to GVariant available outside the module so that
we can consume it for live unwinding.
This allows copying events from a capture stream transparently into the
destination. No processing of the stream is performed, but that may change
in the future to accomidate JIT/Counter translations.
Internal only as support for upcoming live unwinding via external process.
Even if it is disabled, we want the GType enabled and part of our ABI. We
will return NULL if one is created and debuginfod is not supported or if
we failed to create a client.
If we are configured with support for debuginfod and it fails to initialize
or we are not configured to use debuginfod, always ensure g_object_new()
will return a NULL instance back.
This helps prevent against misconfigured instances so we do not need to
do so many checks in vfuncs as well as allowing us to have the GType be
available always even when not built with debuginfod.
This is based on a debuginfod_client provided by Barnabás Pőcze in !73.
It extends it to use the new task infrastructure to elevate the download
process to the user while loading a capture file.
This makes sure that we get that even when not using the Elf symbolizer
but also means we can fallback through the Elf symbolizer into the
debuginfod symbolizer.
This provides a task abstraction to SysprofDocumentLoader so that we can
elevate information about tasks to the user interface. It also moves the
spinner to a menu button w/ popover to display those tasks.
That will cause infinite recursion in `sysprof_elf_get_symbol_at_address_internal()`.
Also note that loops are still possible, this change
only prevents one way of creating loops.
Previously, the `dbus-monitor` process was never terminated.
Unfortunately, tying up some memory is not the worst consequence
of this: after some time file descriptors that are sent on the
bus start to accumulate. This can be especially problematic
when there are unexpected read/write ends of pipes.
It's a simple CLI tool that dumps the contents of a Sysprof capture
file in a syntax inspired by GTK's render node syntax.
It prints metadata, the callgraph, marks, and counters. Other fields
may be added as needed, but for now this is sufficient for inspecting
WebKit captures.
There are CLI args for not printing particular groups, so that we can
filter out what we need.
This is a private allocator that can be used for large groups of
allocations that are tied to a specific object.
For example, the callgraph owning many nodes can use this so all the nodes
are allocated together, but also freed in a single stage ignoring all the
complex GList linked nodes among them.
This simplifies deployment on embedded devices, where polkit is usually
unncessary at runtime, but pulls in quite a few otherwise unncessary
dependencies. Start to improve the situation by allowing to selectively
disable polkit-agent support at compile time, which aids in container
usage scenarios, where one wants to invoke 'sysprof-cli' from within
the container. Bypassing polkit-agent in the container is then desired,
since the host sysprofd will handle asking for permissions to enable
the tracing. It allows for a simpler setup of rootless podman
containers, avoiding UID mismatches, that lead to rejection of the
tracing enablement.
- Add a new 'polkit-agent' meson build feature, that allows to force disabling
polkit-agent support (-Dpolkit-agent=disabled).
- Mark the 'polkit-agent' feature as enabled, by default, to reflect
the current status (sysprof-cli did not build without polkit-agent support).
- libsysprof/sysprof-instrument.c: Build fix when polkit is not available,
remove the unnecessary 'g_autopr(PolkitDetails) details' variable.
- Alter the sysprof-cli dependencies to only attempt to link against
polkit-agent, if necessary. Modify sysprof-cli.c to wrap all code using
polkit-agent in HAVE_POLKIT_AGENT blocks.
