On 2/2/26 9:55 PM, Daniel Hiepler wrote:
Digging into the linuxcncrsh code uncovered quite a few interesting
things (apart from the off-by-one errors and other problems).
Chiming in because I use linuxcncrsh with my own patches: https://
github.com/LinuxCNC/linuxcnc/pull/2739 (got merged but caused the
existing test to fail so it was reverted). I certainly didn't catch
all issues but maybe it's interesting for anyone willing to give
linuxcncrsh some polish.
Thanks for the link!
It needs a bit more than TLC... There are many problems; off-by-one,
race conditions, buffer overruns, static buffers and other non-reentrant
code, possible leaks, etc... And then, the source compiles as a C++
program, but the code is really written as C.
At least I can look at the test(s) and see how good/bad my changes are
so far.
--
Greetings Bertho
(disclaimers are disclaimed)
_______________________________________________
Emc-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-developers
