If a function is bound by a `letrec' (or internal definition)
that cannot be simplified to `let' or lifted to a constant or
top-/module-level function, and if the `letrec' binding is used in
a non-application position, and if the function has in its closure
a typed binding (i.e., boxed, fixnum, flonum, or extflonum),
then the validator was incorrectly rejecting the function's
bytecode --- because the validator didn't distinguish between typed
arguments and typed closure content.
File under "surprised that we didn't hit this one earlier".
Stress mode forces a GC on every N allocation attempts, and it makes
JIT-generated code always take a slow path.
This mode uncovered only a bad test case and some boring start-up
bugs (boring because start-up is deterministic enough that they
never happen), so far.
A package catalog now supplies information about a package's modules
and dependencies, so propagate it when copying a catalog, make the
information accessible via `raco pkg catalog-show', and so on.
The safe-for-space pass could add clearing operations on "typed"
stack positions, which are known to contain a fixnum, flonum, or
extflonum. Non-clearing references, however, were not annotated to
indicate that clearing references were present, since clearing is
not expected on typed positions.
Along the lines of not expecting clearing, the bytecode validator's
encoding of the stack doesn't accomodate both "has a type" and "claims
never to be cleared", so it couldn't detect the bytecode compiler
bug. (Also, this problem didn't show up in the HOSC paper's model of
the validator, because the model pre-dates type tracking.)
Fix the bytecode compiler's space-safety pass so that it never inserts
clearing operations for typed stack positions. Then, the validator can
simply reject any attempt to clear a typed position.
Also, annotate applications generated by lifting as safe-for-space
tail calls.
Merge to v5.3.4
Also, more consistently propagate a given checksum, which can
happen more through the `pkg-install' export from `pkg/lib'
than through `raco pkg'.
Also, report to the user when consulting GitHub or downloading
a checksum.
- build the sub projections only once and build them before getting
the blame objects (instead of after getting the actual values)
- added context information to the blame objects
- added an optional #:min-count argument to insist on a minimum length
for the stream
related to PR 13709
A constant result for foreign-thread use of a callback allows a
callback to return without synchronizing with the Racket thread.
A constant result is thus useful when a callback's work can simply
be skipped if the callback is applied in the "wrong" OS thread.
This termonology change affects lots of function names from `pkg/lib'
and `pkg/db' (former `pkg/pnr-db'), and it also affects some `raco
pkg' commands.
Existing package installations that are marked as 'pnr in a
local configuration are converted automatically to 'catalog, but any
existing "indexes" configuration must be changed to "catalogs".
Bug introduced by 7a8c2ff063: a tree can be deep enough that the
representation of the path to the current item can be too big to
fit into 32 bits. It will always fit in 64 bits, though.
Merge to 5.3.4
