SISC

Settings | Report Duplicate

6

I Use This!

Activity Not Available

Commits : Listings

Analyzed 4 months ago. based on code collected over 1 year ago.

Commit Message	Contributor	Files Modified	Lines Added	Lines Removed	Code Location	Date
Dec 01, 2021 — Dec 01, 2022 Showing page 7 of 107 Search / Filter on:
* added srfi-78 This also resolves FR 1461890.	mradestock	More...				about 18 years ago
* refactored srfi-42 interface Due to the peculiarities of psyntax' module system, any module that exports macros which expand into syntax containing srfi-42 macros needs to export a whole bunch of 'hidden' srfi-42 bindings. To make this less painful we define an interface containing these bindings, thus allowing them to be exported from other modules by defining a compound interface.	mradestock	More...				about 18 years ago
* fixed srfi-55's \|require-extension\| to work in compiled code. See bug 1471254. This is now a macro that calls require-library at expansion time and produces code that calls it at execution time and performs an import. It is because of the latter that we need to expansion-time evaluation since the 'import' cannot be expanded unless the library has been loaded.	mradestock	More...				about 18 years ago
* add list of current and planned srfi support	mradestock	More...				about 18 years ago
* removed srfi-24 we only implement it partially - letrec-mixed is missing - so it's wrong to claim that we support it.	mradestock	More...				about 18 years ago
* corrected typo in hashtable/remove! docs	mradestock	More...				about 18 years ago
* Added some breakpoint helper functions. See FR 1452126. Added functions to inspect and modify breakpoint arguments, and to obtain breakpoint continuations.	mradestock	More...				about 18 years ago
* fixed a dormant bug in breakpoint logic When a breakpoint was hit the repl was being called like this (((getprop 'repl 'debug) (lambda () ...))) which is one set of parentheses too many. This doesn't actually matter though since the call is to a continuation and hence never returns.	mradestock	More...				about 18 years ago
* store break point info on module-local var rather than 'debug env It's cleaner that way.	mradestock	More...				about 18 years ago
* don't display top of stack when hitting break point The top of the stack usually contains the call/cc frame of the breakpoint logic and hence is of limited use.	mradestock	More...				about 18 years ago
* minor code reformatting	mradestock	More...				about 18 years ago
* refactored and tweaked tracing We now use a module-local hashtable to keep track of traced procedures, rather than an alist held in the 'debug env. Furthermore, \|untrace\| now behaves like \|trace\| in the zero arg case, i.e. it displays the list of traced procedures. It also, like \|trace\|, re-establishes tracing for any rebound traced procedures.	mradestock	More...				about 18 years ago
* store breakpoint details in a module-local hashtable this is simpler and cleaner previously we used an alist in the 'debug environment	mradestock	More...				about 18 years ago
* print message telling users how to enabled detailed stack tracing we only print this message once	mradestock	More...				about 18 years ago
* some minor refactoring in debug.scm introduce function to check whether detailed stack tracing is enabled	mradestock	More...				about 18 years ago
Prep for release	scgmille	More...				about 18 years ago
lite changes for compile, javadoc inclusion	scgmille	More...				about 18 years ago
lite changes for compile	scgmille	More...				about 18 years ago
Pass the debugging proc to SRFI 14's lowest-bit-set	scgmille	More...				about 18 years ago
Updated version	scgmille	More...				about 18 years ago
Make sure that corresponding debugging and expression execution happens in the locality its supposed to, so dynamic environments are used correctly. Fixes bug 1462037	scgmille	More...				about 18 years ago
* explain interrupt handling, and don't set nxp to null The latter caused reported stack traces to be empty since a null nxp is a stack trace marker. The error k always skips the top frame, so the nxp can actually be anything, without impact on execution.	mradestock	More...				about 18 years ago
* move trace() call in AppEval.eval() before interrupt handing This doesn't make much difference in practice but is more consistent since it results in the interrupt error being raised after the current application has been recorded in the stack trace, just as it would happen with any other error raised during application.	mradestock	More...				about 18 years ago
* make Interpreter.stk private access to the stk register is now completely contained in Interpreter, following recent changes to k capture implementation	mradestock	More...				about 18 years ago
* fix space leak in \|call-with-values\| This fixes bug 1461939. The problem was that \|call-with-values\| used push(), which resulted in the current vlr ending up in a call frame on the stack. That vlr was then recycled by the subsequent call to setupTailCall(). The solution is to push a frame that has all fields which are critical for safe-for-space set to null.	mradestock	More...				about 18 years ago
* remove lxp Interpreter register This was used in order to keep track of the last executed expression so errors from prims (which all set nxp to null on entry) got sensible location info. However, there is actually a better way to do this - simply restore the original nxp when an error is encountered. The advantage of this, aside from removing an Interpreter register, is that we now also get sensible location info for errors in prims that set the nxp.	mradestock	More...				about 18 years ago
* improve stack traces for captured continuations When we capture a k the current stack tracer may have recorded lots of useful data. In order to not lose that we prepend a dummy frame to the captured call frame stack. Ideally that frame would have a null nxp, so it gets popped automatically. However, frames with null nxps serve as stack trace markers. The dummy frame we create therefore is given a nxp and we wrap it in an ApplyParentFrame in order to ensure that the dummy nxp never gets executed.	mradestock	More...				about 18 years ago
* introduced small helper method on Interpreter to copy the stack tracer	mradestock	More...				about 18 years ago
* fixed space leak in \|with-stack-marker\| This fixes bug 1461425. The problem was that \|with-stack-marker\| used push(), which resulted in the current vlr ending up in a call frame on the stack. That vlr was then recycled by the subsequent call to setupTailCall(). The solution is to push a frame that has all fields which are critical for safe-for-space set to null.	mradestock	More...				about 18 years ago
* a tiny bit of reformatting	mradestock	More...				about 18 years ago

←
1
2
3
4
5
6
7
8
9
10
11
…
106
107
→