The main event loop pushes itself onto the event loop stack, and so it
should also pop itself when destroyed.
This will surface attempts to use the event loop stack after the main
event loop has been destroyed.
For example, consider the following SecretString construction:
String foo = "foo";
auto ss = SecretString::take_ownership(foo.to_byte_buffer());
The ByteBuffer created by to_byte_buffer() will not contain the NUL
terminator. Therefore, the value returned by SecretString::characters
will not be NUL-terminated either.
Currently, the only use of SecretString is to pass its character data to
crypt(), which requires a NUL-terminated string. To ensure this cannot
result in a buffer overrun, make SecretString append a NUL terminator to
its buffer if there isn't one already.
This matches the API of Account::authenticate. The only caller to this
API is the passwd utility, which already has the new password stored as
a SecretString.
When a socket's user doesn't need it to be active, but wants to keep it
open, the socket's notifiers should not be enabled to avoid hogging the
CPU with effectively useless notifications.
This API can be used to disable said notifiers until the user needs the
notifications.
This restriction doesn't make much sense, a user should be free to
consume the buffered data with as small a max_size as they desire.
This fixes a possible RequestServer spin when talking to HTTP servers.
If the server responds with this header, we _must_ close the connection,
as the server is allowed to ignore the socket and not respond to
anything past that response.
Fixes some RequestServer spins.
Prior this change, if you had a program which frequently reads small
amount of bytes, then it would constantly fire up syscalls.
This patch sets that the minimum size that is passed to the read syscall
is 1024 and then it saves these additional bytes in a buffer for next
reads, which greatly improves the cpu usage on such cases.
In other words: reading flacs is now very efficient.
This was used to work around a possible bug where select() would mark a
socket readable, but another user of the same socket would read before
the notifier's callback is run; let's remove this and fix any issues
regarding that specific situation later when they pop up.
This makes connections (particularly TLS-based ones) do the handshaking
stuff only once.
Currently the cache is configured to keep at most two connections evenly
balanced in queue size, and with a grace period of 10s after the last
queued job has finished (after which the connection will be dropped).
We have a few places where we read secrets into memory, and then
do some computation on them. In these cases we should always make
sure we zero the allocations before they are free'd.
The SecureString wrapper provides this abstraction by wrapping a
ByteBuffer and calling explicit_bzero on destruction of the object.
File::link_file takes the dst_path then the src_path so on duplicate
names we tried to create a link at the original file location, which
then flipped the parameters back round again and we ended up with a
broken link from "dst_path (1)" to "src_path (1)".
Command used:
grep -Pirn '(out|warn)ln\((?!["\)]|format,|stderr,|stdout,|output, ")' \
AK Kernel/ Tests/ Userland/
(Plus some manual reviewing.)
Let's pick ArgsParser as an example:
outln(file, m_general_help);
This will fail at runtime if the general help happens to contain braces.
Even if this transformation turns out to be unnecessary in a place or
two, this way the code is "more obviously" correct.
`ArgsParser` and `AboutDialog` had the same procedure to read the
version from `/res/version.ini`. Now they use the `SERENITY_VERSION`
string by default.
This commit refactored the version-reading utility to the new
`Core::Version` namespace.
Before, `AboutDialog` and `ArgsParser` read from a build-time created
file called `/res/version.ini`. This caused problems with utilities
unveiling specific paths leaving the version file unaccessible.
This commit hard-codes a serenity version in `LibCore`, and use it in
`ArgsParser` and `AboutDialog`.
The previous version contained the hash of the last GIT commit, this is
omitted for the default use for the sake of simplicity.
Core::deferred_invoke is a way of executing an action after previously
queued events have been processed. It removes the requirement of
having/being a Core::Object subclass in order to defer invocation
through Core::Object::deferred_invoke.
Core::Object::deferred_invoke now delegates to Core::deferred_invoke.
The version with the Object& argument is still present but will be
removed in the following commits.
This commit additionally fixes a new places where the
DeferredInvocationEvent was dispatched to the event loop directly, and
replaces them with the Core::deferred_invoke equivalent.
We don't need to be allocating Strings for these names during static
initialization. The C-string literals will be stored in the .rodata ELF
section, so they're not going anywhere. We can just wrap the .rodata
storage for the class names in StringViews and use those in Object
registration and lookup APIs.
This static bool getter can be used to VERIFY that an event loop exists,
in situations where one is expected.
This is helpful if the absence of an event loop would generate strange
and/or loud errors that don't immediately point to this as a cause.
This patch adds support for opening a ConfigFile using a file
descriptor rather than trying to open a the file by name directly.
In contrast to the previous implementation, ConfigFile now always keeps
a reference to an open File and does not reopen it for writing.
This requires providing an additional argument to open functions if a
file gets opened based on its name and the user of the api intends to
write to the file in the future.
`--version` always prints "git" for now.
The motivation is that the neofetch port calls `Shell --version` and
adds the output to its output. And if `Shell --version` prints a long
error message about it not knowing the flag, neofetch's output looks a
bit ugly. Per Discord discussion, just add the flag to ArgsParser
instead of only to Shell.
Previously the system would assume the socket was connected after the
file descriptor became writeable. Just because the fd is signaled as
ready for output does not necessarily indicate the socket is connected.
Instead, we should check the status of the socket with SO_ERROR and
handle successes/errors accordingly.
