A real breakpoint makes entries in the breakpoint filter for all offsets at the given line. Various flow control commands will also use the breakpoint filter to add and remove temporary "internal breakpoints" required during the flow operation. Ensure that we never remove a breakpoint filter entry if there was already one there due to a breakpoint (or really any other reason).
When the proxy's signal polling thread gets a signal from the debugger client, it would normally wait one second for another thread to recognize and consume the signal. This is pretty reasonable, but with a debug build on a heavily loaded system it's very rarely possible for one second to not be long enough. Added a runtime option to extend this, and set it to 3s for the existing debugger server tests.
We use ArrayKind as a simple unsigned integer, its an index, we do
range-based comparisons, and in the future we'll likely do bit-ops
for fast sub-kind tests. Just be honest that this is an old-school enum.
The shell builtin implementation of echo on MacOSX
doesn't respect the -n flag. This is the only usage of it
in our stack so far and we can avoid it by hiding the generated
token from the linter in a different way.
Introduces a new m_kind for vector-shaped HphpArray instances.
A vector has integer indexes in the range [0..size), and no
tombstones. Internally it does not store the integer keys or
hashtable, although this version still allocates space for it.
No emitted bytecode relies on Continuation being stored in local 0
anymore. Stop using local 0 for this purpose and compute offset
to the Continuation at JIT time. 16 bytes of memory freed.
At this point all locals of Continuation construction wrapper share the
same indices with their respective locals of Continuation body, which
should allow further optimizations.
The the code is being translated, we statically know offset of
Continuaton from its ActRec. Use this knowledge to emit data fetches
relative to ActRec rather than relative to Continuation loaded from
local 0.
This diff renames the Tracelet -> RegionDesc conversion mode to
"legacy" (since it's going away eventually) and changes "tracelet" to use the
new region selection mode. It attempts to select a region that will be the same
length as what Translator::analyze would come up with, using HhbcTranslator for
all of the type flow logic. It generates longer tracelets in some cases due to
more precise type information. Once this new mode is no longer a perf
regression it can become the new default, replacing all the code in
Translator::analyze and the "legacy" region mode. This version doesn't support
inlining or tracking of known Func*s; those will come in later diffs.
The exit path for the debugger client has always been a little odd. We'd call a shutdown function which would destroy the client, then later call a stop function which would first make a new client, then stop it, then that one would get destroyed later. Made it so we stop and destroy just one client.
Flow control commands should not evaluate the conditions of conditional breakpoints when enabling and disabling breakpoints during stepping. Also, all breakpoints, rather than just the first matching breakpoint should be enabled/disabled.
There was a race in the debugger protocol tests which send bad commands to the proxy and expect it to disconnect. If the proxy thread was fast enough it would mark the client as stopped, and we'd skip trying to get a new message after sending the bad command. This means missing one line of output in very rare cases. I've put in a wait to ensure that the client settles in the stopped state before continuing, to remove the race.
We'd like to start using ##mixed## instead of ##var## for attribute types to be consistent with Hack. As a followup to this (once released), we would codemod all ##var## to ##mixed##.
A client couldn't break execution during eval. There used to be a lot of barriers to making that right, but I fixed most of them with a previous diff on unifying client-side event loops. Now the only barrier was that a server-side thread processing an interrupt was blocking the signal polling thread by holding a mutex while processing the interrupt. Changed to set a flag to disable polling when starting to process the interrupt (and unsetting it when done), while still synchronizing with the signal polling thread to ensure only one thread is sending the client messages at a time. Added logic to re-enable polling while executing PHP for eval, print, etc. Plumbed the proxy thru to the point where we check the clause on conditional breakpoints, too, since that's the third (and final) place we do this.
When the client sends a list of breakpoints over to the server, the logic that applies the breakpoints as soon as a file is loaded is broken. It stops looking at breakpoints as soon as it finds the first one. It should carry on and look at all of them.
This diff creates IRTranslator, which creates and uses an
HhbcTranslator to implement the translate* methods. It can be used
independently of Translator or TranslatorX64 (it isn't yet but my next
region compiler diff uses it). I also moved a bunch of methods out of
inappropriate classes and changed the type guard/assert methods in
HhbcTranslator to use RegionDesc::Location instead of Transl::Location
and fixes a local tracking issue in translateRegion.
"foo" | "f" is meant to yield a string of length
max(strlen("foo"), strlen("f")) by bitwise ORing each byte
of one string against the corresponding byte of the other.
When the strings are of differing lengths, however, we get unknown
garbage data from past the end of the buffer. This data is
often '\0', however under memory pressure we can get any
value and the behavior becomes undefined.
Having everybody and their uncle reading and writing fields out
of a.code and stubs.code was making assembler work hard. This replaces most
reads with accessors, and all writes with structured friends of
X64SAssembler.
This is an incremental step towards moving it all the way
to hphp/server. This flattens base but doesn't untangle
the server files from lib_hphp_runtime
Often I happen not to be running under a debugger when a
crash occurs. In my debugging workflow, I'd prefer our crash handler to
preserve the broken state and spin waiting for gdb to attach rather than
dump core, attempt its own little sad stacktraces, etc. Driven via a
default-false config flag.
I broke the operation of CmdMachine when we attach with the force flag with https://phabricator.fb.com/D851880. I consolidated some error handling logic around broken connections, etc., and in particular pulled throwing of an exception into forceQuit(). I failed to realize there was a single call site of forceQuit() that did not in fact throw an exception afterwards… the logic underneath of force attach. Switching the forceQuit() call there to stop() replaces the logic that was there previously, and restores the behavior.
break start/end/psp currently always report themselves as unbound. If the client is connected to a sanbox, these should instead be treated as bound. Also, break clear all currently removes breakpoints without running their destructors in the right order, which causes the break point counter to not reset to 1.
Fix a long-standing bug where if you step off of the end of pseudo main we'd segfault in the interpreter loop. To get this we have to have a breakpoint in the final TC in pseudo main, and have that called from another TC, then Step or Next off the end. We'd end up with a PC of zero after the ret, which makes sense, but the logic after interpreting a block of code to stay in the interpreter when debugging was blind to this case.
Mike Magruder