diff --git a/CVE-2025-6069-quad-complex-HTMLParser.patch b/CVE-2025-6069-quad-complex-HTMLParser.patch new file mode 100644 index 0000000..70eb744 --- /dev/null +++ b/CVE-2025-6069-quad-complex-HTMLParser.patch @@ -0,0 +1,237 @@ +From 1d53c3e7343bddb064182e02c21b13be9b63390f Mon Sep 17 00:00:00 2001 +From: Serhiy Storchaka +Date: Fri, 13 Jun 2025 19:57:48 +0300 +Subject: [PATCH] [3.12] gh-135462: Fix quadratic complexity in processing + special input in HTMLParser (GH-135464) + +End-of-file errors are now handled according to the HTML5 specs -- +comments and declarations are automatically closed, tags are ignored. +(cherry picked from commit 6eb6c5dbfb528bd07d77b60fd71fd05d81d45c41) + +Co-authored-by: Serhiy Storchaka +--- + Lib/html/parser.py | 41 +++- + Lib/test/test_htmlparser.py | 94 ++++++++-- + Misc/NEWS.d/next/Security/2025-06-13-15-55-22.gh-issue-135462.KBeJpc.rst | 4 + 3 files changed, 116 insertions(+), 23 deletions(-) + create mode 100644 Misc/NEWS.d/next/Security/2025-06-13-15-55-22.gh-issue-135462.KBeJpc.rst + +Index: Python-3.12.11/Lib/html/parser.py +=================================================================== +--- Python-3.12.11.orig/Lib/html/parser.py 2025-07-02 17:09:00.904899297 +0200 ++++ Python-3.12.11/Lib/html/parser.py 2025-07-02 17:09:12.496469955 +0200 +@@ -25,6 +25,7 @@ + charref = re.compile('&#(?:[0-9]+|[xX][0-9a-fA-F]+)[^0-9a-fA-F]') + + starttagopen = re.compile('<[a-zA-Z]') ++endtagopen = re.compile('') + commentclose = re.compile(r'--\s*>') + # Note: +@@ -177,7 +178,7 @@ + k = self.parse_pi(i) + elif startswith("', i + 1) +- if k < 0: +- k = rawdata.find('<', i + 1) +- if k < 0: +- k = i + 1 ++ if starttagopen.match(rawdata, i): # < + letter ++ pass ++ elif startswith("'), +- ('comment', '/img'), +- ('endtag', 'html<')]) ++ ('data', '\n')]) + + def test_starttag_junk_chars(self): ++ self._run_check("<", [('data', '<')]) ++ self._run_check("<>", [('data', '<>')]) ++ self._run_check("< >", [('data', '< >')]) ++ self._run_check("< ", [('data', '< ')]) + self._run_check("", []) ++ self._run_check("<$>", [('data', '<$>')]) + self._run_check("", [('comment', '$')]) + self._run_check("", [('endtag', 'a')]) ++ self._run_check("", [('starttag', 'a", [('endtag', 'a'", [('data', "'", []) ++ self._run_check("", [('starttag', 'a$b', [])]) + self._run_check("", [('startendtag', 'a$b', [])]) + self._run_check("", [('starttag', 'a$b', [])]) + self._run_check("", [('startendtag', 'a$b', [])]) ++ self._run_check("", [('endtag', 'a$b')]) + + def test_slashes_in_starttag(self): + self._run_check('', [('startendtag', 'a', [('foo', 'var')])]) +@@ -539,13 +546,56 @@ + for html, expected in data: + self._run_check(html, expected) + +- def test_broken_comments(self): +- html = ('' ++ def test_eof_in_comments(self): ++ data = [ ++ ('', [('comment', '-!>')]), ++ ('' + '' + '' + '') + expected = [ ++ ('comment', 'ELEMENT br EMPTY'), + ('comment', ' not really a comment '), + ('comment', ' not a comment either --'), + ('comment', ' -- close enough --'), +@@ -600,6 +650,26 @@ + ('endtag', 'a'), ('data', ' bar & baz')] + ) + ++ @support.requires_resource('cpu') ++ def test_eof_no_quadratic_complexity(self): ++ # Each of these examples used to take about an hour. ++ # Now they take a fraction of a second. ++ def check(source): ++ parser = html.parser.HTMLParser() ++ parser.feed(source) ++ parser.close() ++ n = 120_000 ++ check(" +Date: Tue, 1 Jul 2025 12:13:28 +0200 +Subject: [PATCH] Doc: Generate ids for audit_events using docname + +This patch generates ids for audit_events using the docname so the id is +not global but depend on the source file. This make the doc build +reproducible with multiple cores because it doesn't which file is parsed +first, the id for audit_events will always be consistent independently +of what file is parsed first. + +https://github.com/python/cpython/issues/130979 +--- + Doc/tools/extensions/audit_events.py | 11 ++++++++--- + 1 file changed, 8 insertions(+), 3 deletions(-) + +Index: Python-3.12.11/Doc/tools/extensions/audit_events.py +=================================================================== +--- Python-3.12.11.orig/Doc/tools/extensions/audit_events.py 2025-07-02 16:10:55.447792590 +0200 ++++ Python-3.12.11/Doc/tools/extensions/audit_events.py 2025-07-02 16:12:00.124596479 +0200 +@@ -64,8 +64,13 @@ + logger.warning(msg) + return + +- def id_for(self, name) -> str: +- source_count = len(self.sources.get(name, ())) ++ def _source_count(self, name, docname) -> int: ++ """Count the event name in the same source""" ++ sources = self.sources.get(name, set()) ++ return len([s for s, t in sources if s == docname]) ++ ++ def id_for(self, name, docname) -> str: ++ source_count = self._source_count(name, docname) + name_clean = re.sub(r"\W", "_", name) + return f"audit_event_{name_clean}_{source_count}" + +@@ -140,7 +145,7 @@ + except (IndexError, TypeError): + target = None + if not target: +- target = self.env.audit_events.id_for(name) ++ target = self.env.audit_events.id_for(name, self.env.docname) + ids.append(target) + self.env.audit_events.add_event(name, args, (self.env.docname, target)) + diff --git a/python312.changes b/python312.changes index 4173fd7..3d0a360 100644 --- a/python312.changes +++ b/python312.changes @@ -1,3 +1,24 @@ +------------------------------------------------------------------- +Wed Jul 2 14:47:20 UTC 2025 - Matej Cepl + +- Add CVE-2025-6069-quad-complex-HTMLParser.patch to avoid worst + case quadratic complexity when processing certain crafted + malformed inputs with HTMLParser (CVE-2025-6069, bsc#1244705). + +------------------------------------------------------------------- +Wed Jul 2 13:14:28 UTC 2025 - Matej Cepl + +- Add bsc1243155-sphinx-non-determinism.patch (bsc#1243155) to + generate ids for audit_events using docname (reproducible + builds). + +------------------------------------------------------------------- +Tue Jul 1 08:19:52 UTC 2025 - Daniel Garcia + +- Use one core to build doc. This will make sphinx doc build + reproducible. + bsc#1243155 + ------------------------------------------------------------------- Mon Jun 9 19:41:07 UTC 2025 - Matej Cepl diff --git a/python312.spec b/python312.spec index 17fba23..5caacd0 100644 --- a/python312.spec +++ b/python312.spec @@ -184,6 +184,12 @@ Patch41: docs-docutils_014-Sphinx_420.patch # PATCH-FIX-SLE doc-py38-to-py36.patch mcepl@suse.com # Make documentation extensions working with Python 3.6 Patch44: doc-py38-to-py36.patch +# PATCH-FIX-UPSTREAM bsc1243155-sphinx-non-determinism.patch bsc#1243155 mcepl@suse.com +# Doc: Generate ids for audit_events using docname +Patch45: bsc1243155-sphinx-non-determinism.patch +# PATCH-FIX-UPSTREAM CVE-2025-6069-quad-complex-HTMLParser.patch bsc#1244705 mcepl@suse.com +# avoid quadratic complexity when processing malformed inputs with HTMLParser +Patch46: CVE-2025-6069-quad-complex-HTMLParser.patch BuildRequires: autoconf-archive BuildRequires: automake BuildRequires: fdupes @@ -486,7 +492,7 @@ TODAY_DATE=`date -r %{SOURCE0} "+%%B %%d, %%Y"` cd Doc sed -i "s/^today = .*/today = '$TODAY_DATE'/" conf.py -%make_build -j1 html +%make_build -j1 JOBS=1 html # Build also devhelp files sphinx-build -a -b devhelp . build/devhelp