summaryrefslogtreecommitdiff
path: root/Src
AgeCommit message (Collapse)AuthorFilesLines
2015-01-0634108: Don't leak ifs stuffMikael Magnusson1-2/+2
Found by Coverity (Issue 1255785).
2015-01-0634107: getsubsargs: free ptr1 before returningMikael Magnusson1-0/+1
Found by Coverity (Issue 439073).
2015-01-0634134: anon funcs: don't leak shf and related dataMikael Magnusson1-0/+6
Found by Coverity (Issue 439076).
2015-01-0634112: typeset: fix leak of oldvalMikael Magnusson1-0/+2
Found by Coverity (Issue 1255803).
2015-01-0634106: hist: use zhtricat instead of tricatMikael Magnusson1-1/+1
Found by Coverity (Issue 1255769).
2015-01-0634113: whence: use dupstring to not leak memoryMikael Magnusson1-1/+1
All other assignments to buf use the heap, and it's never freed. Found by Coverity (Issue 1255786).
2015-01-0634119: complist: Fix leak of string in clnicezputsMikael Magnusson1-0/+2
Found by Coverity (Issue 1255808).
2015-01-0634105: subst: remove dead codeMikael Magnusson1-1/+1
Found by Coverity (Issue 1255810).
2015-01-0634121: compresult: Remove unneeded NULL checkMikael Magnusson1-1/+1
The variable is set to if NULL at the start of the function, and derefed on the previous line. Found by Coverity (Issue 1255843).
2015-01-0634104: compctl: Remove pointless checkMikael Magnusson1-1/+1
cc has already been derefed a bunch of times leading up to here. Found by Coverity (Issue 1255841).
2015-01-0634115: compcore: Fix size argument to zfreeMikael Magnusson1-1/+1
Found by Coverity (Issue 1255852), has no impact unless using --enable-zsh-mem, and even then it is minimal.
2015-01-0634117: zle: size_t is unsigned, use int insteadMikael Magnusson1-1/+1
The function wctomb returns an int according to my manpage, and we furthermore check if it is negative, and then return it, and the function signature is int, so declaring it as an int seems to make more sense.
2015-01-0634116: computil: Check for NULL before passing to strlenMikael Magnusson1-1/+2
The rest of this function appears to be very careful about checking these, then forgets in this one spot. Found by Coverity (Issue 1255805).
2015-01-0634118: Don't crash when writing out history if HOST is unsetMikael Magnusson1-1/+2
Found by Coverity (Issue 1255793).
2015-01-0634114: emulate: Handle aborting from mixed -L/-c correctlyMikael Magnusson1-1/+2
Somehow Coverity found this (Issue 1255797, Failure to restore non-local value).
2015-01-0634103: fix ancient double-quote handling thinko in subst_parse_str()Barton E. Schaefer1-2/+4
This doesn't seem to have mattered, but must in some obscure cases
2015-01-0634093: "whence" should always return nonzero when it finds that nothing ↵Barton E. Schaefer1-15/+17
matches its arguments
2015-01-0434092: fix miscount of symlink resolution for "..".Peter Stephenson1-0/+2
This caused problems with expanding a path with ".." in "whence -S".
2015-01-0434091: typo with "whence -s" expansionsPeter Stephenson1-2/+2
2015-01-02users/19671: remove confusion with whence -a.Peter Stephenson1-1/+1
If the argument is a full path don't try to search the path for it.
2015-01-02users/19667: whence -S shows intermediate steps in symlink expansionPeter Stephenson2-13/+66
2014-12-2834070: fix starting position for memset() from 34005.Daniel Shahaf1-1/+1
2014-12-2734064: assignment before command replaces array with export even when KSH_ARRAYSTakeshi Banse1-0/+2
2014-12-2734065: following an "if" condition, do not test lastval for ERR_EXIT until a ↵Barton E. Schaefer2-5/+15
new command is run Includes unposted regression tests.
2014-12-1934015: disallow strange environment variable names.Peter Stephenson1-16/+23
These are ones with the top bit set in any character. Don't import them, and don't export them.
2014-12-1934005: region_highlights memory fixPeter Stephenson1-2/+8
Zero uninitialised part of memory when reallocing
2014-12-1834002: zshcalloc() in init_keymaps()Barton E. Schaefer1-1/+1
2014-12-1833992: do not attempt attachtty() for process group zero (which is possible ↵Barton E. Schaefer1-2/+5
in a linux pid namespace)
2014-12-1834008: metafy the environment on arrival in the shellPeter Stephenson1-1/+1
2014-12-1834006: unmetafy anything put into the environmentPeter Stephenson1-1/+11
2014-12-1833981: more care with region_highlights managementPeter Stephenson2-29/+39
2014-12-1833978: avoid infinite loop in interactive modeJun-ichi Takimoto1-1/+3
Update command line when accept-and-hold is called in the interactive mode of menu select.
2014-12-1633984: bin_dirs() should use zputs() to print metafied directory namesJun Kuriyama1-1/+1
2014-12-1633982: minimal support for pid namespaces by recognizing that GETPGRP() may ↵Chirantan Ekbote1-1/+1
return 0
2014-12-1533976: fix overlapping strcpy()Barton E. Schaefer1-1/+1
2014-12-1333956: document key binding changes and remove ^X bindingOliver Kiddle2-1/+2
2014-12-1333846: additional default vi-mode key bindingsOliver Kiddle3-5/+12
2014-12-1233950: ignore KEYTIMEOUT for vi operatorsOliver Kiddle3-12/+19
2014-12-1133876: etc.: Separate errors and keyboards interruptsPeter Stephenson28-124/+273
Combination of 12 commits from interrupt_abort branch. Basic strategy is to introduce bits to errflag and to set and reset them separately. Remove interrupt status on return to main keymap. Turn off ERRFLAG_INT for always block. Restore bit thereafter: we probably need a new variable in order to allow user interrupts to be reset in the always block. Add TRY_BLOCK_INTERRUPT This works the same as TRY_BLOCK_ERROR, but for a SIGINT, too. Ensure propagation of SIGINT from exited job. If received by foreground job, shell uses ERRFLAG_INT, not ERRFLAG_ERROR, to set the new state. Reset errflag before precmd() Add always block in _main_completion to fix ZLS_COLORS Ensures we get the right state of $ZLS_COLORS at the end of _main_complete even if there's an interrupt. However, the "right state" is a bit messy as it depends on styles.
2014-12-1033940: handle backslash-meta correctly in lexerJun-ichi Takimoto1-1/+13
2014-12-0933932: revise boundary check in unmeta()Jun-ichi Takimoto1-5/+3
2014-12-0833924: allow vi line/characterwise mode to be forcedOliver Kiddle4-3/+26
2014-12-0733894: boundary conditions in unmeta(), unmetafy()Barton E. Schaefer1-3/+5
Check that we aren't running off the end of the string when converting the next byte after a Meta byte. This is just defensive programming in case of bad metafied strings coming through from gettokstr(), some repairs there are likely still needed.
2014-12-0733854: errors end recursion through scanner()Barton E. Schaefer1-1/+1
Makes **/ more easily interruptible
2014-12-0533860: minor fixes to vi mode changesOliver Kiddle4-44/+52
2014-12-0533845: save correct cursor position with push-line from vi command modeOliver Kiddle1-0/+2
2014-11-3033823: overwrite mode shouldn't replace newlinesOliver Kiddle1-11/+10
2014-11-3033820: detect support for realpath() with a NULLOliver Kiddle1-25/+22
argument and fixes to the use of it
2014-11-3033800: remove old workaround for ancient systems toOliver Kiddle1-41/+2
consume typeahead before setting up the terminal
2014-11-3033818: fix types passed to sizeof detected by coverity as being wrongOliver Kiddle5-9/+9