The test mentioned in the subject triggers a segfault. This is with
xapian-core-1.4.16 on OpenBSD. Here is the backtrace:
#0 0x05c9d491bec2 in Xapian::TermIterator::decref() () from
/usr/local/lib/libxapian.so.5.1
#1 0x05c948827062 in Xapian::TermIterator::~TermIterator
This test extracts values from a (key,value) map where multiple entries
can have the same key, and the entries are sorted by key, but not by
value. The test incorrectly assumes that the values will be sorted as
well, so sort the output.
---
test/T610-message-property.sh | 12
1 file
the "prefix" test in T610-message-property extracts values from a
(key,value) map where multiple entries can have the same key, and the
entries are sorted by key, but not by value. The test incorrectly
assumes that the values will be sorted as well, so correct this by
splitting the output using
On Tue, Apr 14, 2020 at 05:38:32PM -0300, David Bremner wrote:
> Olivier Taïbi writes:
> > PS: out of curiosity, why bother with the --gzip feature in notmuch dump
> > and restore when the user can simply pipe to/from a gzip/bzip2/xz/...
> > command?
>
> I beli
As suggested by David Bremner in
https://notmuchmail.org/pipermail/notmuch/2020/029288.html
here is the patch for bug #3: after gzgets() returns NULL (meaning EOF
or error), the error code Z_STREAM_END means EOF and not error.
Context: I am compiling notmuch on OpenBSD which has a rather old zlib
As suggested by David Bremner in
https://notmuchmail.org/pipermail/notmuch/2020/029288.html
here is a separate patch for bug #2: calling gzerror() (indirectly via
gzerror_str()) after gzclose_r is a use after free, according to zlib's manual.
diff --git a/notmuch-restore.c b/notmuch-restore.c
the following diff addresses 3 zlib-related bugs in notmuch.
1) the second argument of gzerror() cannot be NULL, so replace it by a dummy
2) gzerror() cannot be closed after gzclosed(), so just print the error value
instead.
3) in gz_getline(), if gz_error sets its second argument to
