Doing this causes problems when running ert tests, for
instance (Bug#24778). The call to `debug` when `debug-on-error' is
non-nil was introduced in 2015-02-14 "* lisp/emacs-lisp/cl*.el: Use
define-inline and move some code...".
* lisp/emacs-lisp/cl-preloaded.el (cl--assertion-failed): Don't call
`debug' directly.
* nt/README.W32: Describe 64 and 32-bit downloads, optional
dependencies bundle. Remove old material on sourcing dependencies.
Remove references to Windows 95. Update GUI references to recent
Windows. Remove references to Usenet.
* lisp/vc/vc.el (vc-responsible-backend): Clarify that symlinks
are not resolved when the VC backend is reported.
* doc/lispref/files.texi (Truenames): Document
'vc-responsible-backend'. (Bug#23436)
* doc/emacs/maintaining.texi (Version Control Systems): Fix a
typo.
* lisp/simple.el (transpose-sexps): Clarify the conditions for
transposing sexps that are lists or strings. Mention the effect
of the prefix argument. (Bug#24860)
* doc/lispref/modes.texi (Customizing Keywords): Clarify the
'keywords' argument of 'font-lock-remove-keywords'. Suggested by
Hong Xu <hong@topbug.net>. (Bug#24830)
* doc/misc/eudc.texi (Inline Query Expansion): Fix the default value of
'eudc-inline-expansion-format'. (Bug#24840)
Backport:
(cherry picked from commit 1fef1387c3)
With Python 3.5, (python-shell-completion-native-get-completions ... "")
would return an empty list, causing python.el to think that native
completion was unavailable (the difference between Python 2 and Python 3
is due to https://bugs.python.org/issue25660).
* lisp/progmodes/python.el (python-shell-completion-native-try): Use "_"
to check whether native completion is available instead of "".
* src/search.c (looking_at_1, fast_looking_at, search_buffer): Prevent
relocation of buffer contents during calls to re_search_2. This ensures
the pointers into buffer text won't be invalidated by
r_alloc_sbrk (called from malloc with configurations where
REL_ALLOC=yes).
The fix was not complete, and completing it was proving too complicated.
- Revert "* src/regex.c (re_search_2): Make new code safe for
-Wjump-misses-init."
This reverts commit c2a17924a5.
- Revert "Port to GCC 6.2.1 + --enable-gcc-warnings"
This reverts commit f6134bbda2.
- Revert "Fix handling of allocation in regex matching"
This reverts commit ad66b3fadb.
- Revert "Fix handling of buffer relocation in regex.c functions"
This reverts commit ee04aedc72.
This should make ralloc-related bugs less likely on GNU/Linux
systems with bleeding-edge glibc. See the email thread containing:
http://lists.gnu.org/archive/html/emacs-devel/2016-10/msg00801.html
Do not merge to master.
* configure.ac (REL_ALLOC): Default to 'no' on all platforms, not
merely on platforms with Doug Lea malloc. Although bleeding-edge
glibc no longer exports __malloc_initialize_hook and so longer
passes the configure-time test for Doug Lea malloc, ralloc tickles
longstanding bugs like Bug#24358 and Bug#24764 and Emacs is likely
to be more reliable without it. This patch is not needed on
master, which uses hybrid malloc in this situation.
* src/search.c (Freplace_match): Move the call to BYTE_POS_ADDR
after the call to xpalloc, to avoid the danger of buffer text
relocation after its address was taken. (Bug#24358)
* lisp/net/tramp-sh.el (tramp-histfile-override): Change default value
to "~/.tramp_history".
(tramp-open-shell): Check proper HISTFILE setting.
(tramp-maybe-open-connection): Cleanup also for errors. (Bug#24478)
Bleeding-edge glibc sets emacs_cv_var_doug_lea_malloc to 'no'.
Do not merge to master.
* configure.ac: Check for valloc decl when compiling gmalloc.c.
* src/gmalloc.c (emacs_abort) [emacs]: Adjust decl to match
what is in lisp.h. Remove duplicate decl.
(aligned_alloc): #undef before defining.
(aligned_alloc, memalign) [!MSDOS]: Declare.
(valloc) [HAVE_DECL_VALLOC]: Remove duplicate decl.
* src/regex.c (ENSURE_FAIL_STACK, re_search_2):
Redo recent regex changes to avoid complaints from GCC 6.2.1 when
Emacs is configured with --enable-gcc-warnings. Also, work around
GCC bug 78081, which was uncovered by this new code.
Since commit "* lisp/help-fns.el (describe-function): More type
checking[...]", `describe-function' throws a user-error when given a
non-symbol. This prevents the [back] button in a *Help* buffer from
working when the page it goes back to describes an anonymous
function (e.g., the result of `describe-key' on a key which is bound to
a lambda form).
* lisp/help-fns.el (describe-function): Move the checks on FUNCTION
being an fbound symbol into the `interactive' form. This allows
non-interactive calls to pass an anonymous function (Bug #24221). Note
that passing a non-bound symbol non-interactively will still trigger a
`void-function' error from `describe-function-1'.
`re_match_2_internal' uses pointers to the lisp objects that it
searches. Since it may call malloc when growing the "fail stack", these
pointers may be invalidated while searching, resulting in memory
curruption (Bug #24358).
To fix this, we check the pointer that the lisp object (as specified by
re_match_object) points to before and after growing the stack, and
update existing pointers accordingly.
* src/regex.c (STR_BASE_PTR): New macro.
(ENSURE_FAIL_STACK, re_search_2): Use it to convert pointers into
offsets before possible malloc call, and back into pointers again
afterwards.
(POS_AS_IN_BUFFER): Add explanatory comment about punning trick.
* src/search.c (search_buffer): Instead of storing search location as
pointers, store them as pointers and recompute the corresponding address
for each call to `re_search_2'.
(string_match_1, fast_string_match_internal, fast_looking_at):
* src/dired.c (directory_files_internal): Set `re_match_object' to Qnil
after calling `re_search' or `re_match_2'.
* src/regex.h (re_match_object): Mention new usage in commentary.
* doc/lispref/lists.texi (Association Lists): Fix the signature of
'alist-get'. Fix the markup, the wording, and the punctuation in
the description. (Bug#24740)
