Instead of keeping AnonymousVMObject::m_cow_map in an OwnPtr<Bitmap>,
just make the Bitmap a regular value member. This increases the size
of the VMObject by 8 bytes, but removes some of the kmalloc/kfree spam
incurred by sys$fork().
This commit removes the only 3rd party library (and its usages)
in serenity: puff, which is used for deflate decompression. and
replaces it with the existing original serenity implementation
in LibCompress. :^)
This is a little bit messy, but basically if an ELF object is non-PIE,
we have to account for the executable mapping being at a hard-coded
offset and subtract that when doing symbolication.
There's probably a nicer way to solve this, I just hacked this together
so we can see "cc1plus" and friends in profiles. :^)
The QEMU's `--accel hvf` command was recently enabled in the `run.sh`
script, but it sadly doesn't work on macOS Big Sur: you need to first
sign your code by adding an `entitlements.xml` file and running a
simple command.
Since we know for sure that the virtual memory regions in the new
process being created are not being used on any CPU, there's no need
to do TLB flushes for every mapped page.
In multi-process profiles, the same ELF objects tend to occur many
times (everyone has libc.so for example) so we will quickly run out
of VM if we map each object once per process that uses it.
Fix this by adding a "mapped object cache" that maps the path of
an ELF object to a cached memory mapping and wrapping ELF::Image.
Extracted a method from the code in the File Manager application which
added actions for activating launch handlers found for the selected
file from the context menu. Applied this method to desktop files
and shortcuts.
Note: made some launch handler related methods in the DirectoryView
static or const which allows passing const DirectoryView& to certain
methods.
Dynamic Vector allocations in sys$select() were showing up in the
full-system profile and since there will never be more than FD_SETSIZE
file descriptors to worry about, we can confidently add enough inline
capacity to this Vector that it never has to kmalloc.
To compensate for the increased stack usage, reduce the size of the
FDInfo struct while we're here. :^)
The full system profiling functionality is useful for profiling the
boot performance of the system. Add a new kernel boot option to start
the system with profiling enabled. This lets you disable and view a
profile once the system is booted.
You can use it by running:
```
$ run.sh qcmd boot_prof
```
Previously all of the CommandLine parsing was spread out around the
Kernel. Instead move it all into the Kernel CommandLine class, and
expose a strongly typed API for querying the state of options.
Previously, the instruction fetch flag of the page fault handler
did not have the currect binary representation, and would always
return false. This aligns these flags.
Because the z-order changes we not only need to recompute occlusions
but we also need to mark the screen area of the previous and new
highlighted window as dirty.
Fixes#5599
The perfcore file format was previously limited to a single process
since the pid/executable/regions data was top-level in the JSON.
This patch moves the process-specific data into a top-level array
named "processes" and we now add entries for each process that has
been sampled during the profile run.
This makes it possible to see samples from multiple threads when
viewing a perfcore file with Profiler. This is extremely cool! :^)
The superuser can now call sys$profiling_enable() with PID -1 to enable
profiling of all running threads in the system. The perf events are
collected in a global PerformanceEventBuffer (currently 32 MiB in size.)
The events can be accessed via /proc/profile
If we can't allocate a PerformanceEventBuffer to store the profiling
events, we now fail sys$profiling_enable() and sys$perf_event()
with ENOMEM instead of carrying on with a broken buffer.
When you reset() a Track, you need to set the piano roll iterators back
to the first notes.
Fixes#2578. The bug was due to pressing export between 2 notes - the
tracks were never told to go back to the first note.
This was the original approach before we switched to get_fast_random()
which wasn't fast enough, so we added a buffer.
Unfortunately that buffer is racy and we can actually skid past the end
of it and continue fetching "random" offsets from the adjacent memory
for a while, until we run out of kernel data segment and trip a fault.
Instead of making this even more convoluted, let's just go back to the
pleasantly simple (RDTSC & 0xff) approach. :^)
Fixes#4912.
When a mousewheel scroll event isn't handled by the web content
itself (e.g. an overflowed box or similar), the event needs to get
passed back up to the OutOfProcessWebView.
We try scrolling a Node with the handle_mousewheel event, but if it
isn't scrollable, the event should be passed back up to the page
host. This is the first step in that process.
