diff -up thunderbird-91.7.0/parser/expat/lib/xmltok.c.expat-CVE-2022-25235 thunderbird-91.7.0/parser/expat/lib/xmltok.c --- thunderbird-91.7.0/parser/expat/lib/xmltok.c.expat-CVE-2022-25235 2022-03-02 17:57:38.364361168 +0100 +++ thunderbird-91.7.0/parser/expat/lib/xmltok.c 2022-03-02 17:58:22.235512399 +0100 @@ -65,13 +65,6 @@ + ((((byte)[2]) >> 5) & 1)] \ & (1u << (((byte)[2]) & 0x1F))) -#define UTF8_GET_NAMING(pages, p, n) \ - ((n) == 2 \ - ? UTF8_GET_NAMING2(pages, (const unsigned char *)(p)) \ - : ((n) == 3 \ - ? UTF8_GET_NAMING3(pages, (const unsigned char *)(p)) \ - : 0)) - /* Detection of invalid UTF-8 sequences is based on Table 3.1B of Unicode 3.2: http://www.unicode.org/unicode/reports/tr28/ with the additional restriction of not allowing the Unicode diff -up thunderbird-91.7.0/parser/expat/lib/xmltok_impl.c.expat-CVE-2022-25235 thunderbird-91.7.0/parser/expat/lib/xmltok_impl.c --- thunderbird-91.7.0/parser/expat/lib/xmltok_impl.c.expat-CVE-2022-25235 2022-03-02 17:57:38.365361172 +0100 +++ thunderbird-91.7.0/parser/expat/lib/xmltok_impl.c 2022-03-02 18:04:51.240853247 +0100 @@ -34,7 +34,7 @@ case BT_LEAD ## n: \ if (end - ptr < n) \ return XML_TOK_PARTIAL_CHAR; \ - if (!IS_NAME_CHAR(enc, ptr, n)) { \ + if (IS_INVALID_CHAR(enc, ptr, n) || ! IS_NAME_CHAR(enc, ptr, n)) { \ *nextTokPtr = ptr; \ return XML_TOK_INVALID; \ } \ @@ -62,7 +62,7 @@ case BT_LEAD ## n: \ if (end - ptr < n) \ return XML_TOK_PARTIAL_CHAR; \ - if (!IS_NMSTRT_CHAR(enc, ptr, n)) { \ + if (IS_INVALID_CHAR(enc, ptr, n) || ! IS_NMSTRT_CHAR(enc, ptr, n)) { \ *nextTokPtr = ptr; \ return XML_TOK_INVALID; \ } \ @@ -1090,6 +1090,10 @@ PREFIX(prologTok)(const ENCODING *enc, c case BT_LEAD ## n: \ if (end - ptr < n) \ return XML_TOK_PARTIAL_CHAR; \ + if (IS_INVALID_CHAR(enc, ptr, n)) { \ + *nextTokPtr = ptr; \ + return XML_TOK_INVALID; \ + } \ if (IS_NMSTRT_CHAR(enc, ptr, n)) { \ ptr += n; \ tok = XML_TOK_NAME; \