Microsoft DirectWrite / AFDKO – Heap-Based Buffer Overflow in OpenType Font Handling in readCharset
-----=====[ Background ]=====----- AFDKO (Adobe Font Development Kit for OpenType) is a set of tools for examining, modifying and building fonts. The core part of this toolset is a font handling library written in C, which provides interfaces for reading and writing Type 1, OpenType, TrueType (to some extent) and several other font formats. While the library existed as early as 2000, it was open-sourced by Adobe in 2014 on GitHub [1, 2], and is still actively developed. The font parsing code can be generally found under afdko/c/public/lib/source/*read/*.c in the project directory tree. At the time of this writing, based on the available source code, we conclude that AFDKO was originally developed to only process valid, well-formatted font files. It contains very few to no sanity checks of the input data, which makes it susceptible to memory corruption issues (e.g. buffer overflows) and other memory safety problems, if the input file doesn't conform to the format specification. We have recently discovered that starting with Windows 10 1709 (Fall Creators Update, released in October 2017), Microsoft's DirectWrite library [3] includes parts of AFDKO, and specifically the modules for reading and writing OpenType/CFF fonts (internally called cfr/cfw). The code is reachable through dwrite!AdobeCFF2Snapshot, called by methods of the FontInstancer class, called by dwrite!DWriteFontFace::CreateInstancedStream and dwrite!DWriteFactory::CreateInstancedStream. This strongly indicates that the code is used for instancing the relatively new variable fonts [4], i.e. building a single instance of a variable font with a specific set of attributes. The CreateInstancedStream method is not a member of a public COM interface, but we have found that it is called by d2d1!dxc::TextConvertor::InstanceFontResources, which led us to find out that it can be reached through the Direct2D printing interface. It is unclear if there are other ways to trigger the font instancing functionality. One example of a client application which uses Direct2D printing is Microsoft Edge. If a user opens a specially crafted website with an embedded OpenType variable font and decides to print it (to PDF, XPS, or another physical or virtual printer), the AFDKO code will execute with the attacker's font file as input. Below is a description of one such security vulnerability in Adobe's library exploitable through the Edge web browser. -----=====[ Description ]=====----- The readCharset() function in afdko/c/public/lib/source/cffread/cffread.c is designed to read and parse the charset information of an input OpenType font. It is called by cfrBegFont(), the standard entry point function for the "cfr" (CFF Reader) module of AFDKO. The relevant part of the function is shown below: --- cut --- [...] 2179 case 1: 2180 size = 1; 2181 /* Fall through */ 2182 case 2: 2183 while (gid < h->glyphs.cnt) { 2184 unsigned short id = read2(h); 2185 long nLeft = readN(h, size); 2186 while (nLeft-- >= 0) 2187 addID(h, gid++, id++); 2188 } 2189 break; --- cut --- whereas addID() is defined as follows: --- cut --- 1838 /* Add SID/CID to charset */ 1839 static void addID(cfrCtx h, long gid, unsigned short id) { 1840 abfGlyphInfo *info = &h->glyphs.array[gid]; 1841 if (h->flags & CID_FONT) 1842 /* Save CID */ 1843 info->cid = id; 1844 else { 1845 /* Save SID */ 1846 info->gname.impl = id; 1847 info->gname.ptr = sid2str(h, id); 1848 1849 /* Non-CID font so select FD[0] */ 1850 info->iFD = 0; [...] 1859 } 1860 } --- cut --- The readCharset() routine doesn't consider the size of the h->glyphs array and writes to it solely based on the charset information. If the value read from the input stream in line 2185 exceeds the number of glyphs in the font, the array may be overflown in addID() (line 1843 or 1846, 1847, 1850), corrupting adjacent objects on the heap. The h->glyphs array is initialized in readCharStringsINDEX() according to the number of CharStrings found in the font: --- cut --- 1791 dnaSET_CNT(h->glyphs, index.count); --- cut --- -----=====[ Proof of Concept ]=====----- The proof of concept font contains a charset descriptor with the following initial values: - width = 0x02 (changed from 0x01) - id = 0x4141 (changed from 0x0001) - nLeft = 0xffff (changed from 0x15) By increasing the value of "nLeft" from 21 to 65535, we cause the loop in lines 2386-2387 to go largely out of bounds and overflow the h->glyphs array. In theory, the vulnerability shouldn't be possible to reach in Microsoft DirectWrite and its client applications, because AFDKO is only used there for instancing variable fonts, whereas such CFF2 fonts follow another execution path in the readCharset() function: --- cut --- 2138 if (h->header.major == 2) { 2139 postRead(h); 2140 if (h->cff2.mvar) 2141 MVARread(h); 2142 if (!(h->flags & CID_FONT)) 2143 readCharSetFromPost(h); 2144 else { 2145 long gid; 2146 for (gid = 0; gid < h->glyphs.cnt; gid++) { 2147 abfGlyphInfo *info = &h->glyphs.array[gid]; 2148 info->cid = (unsigned short)gid; 2149 } 2150 } 2151 return; 2152 } --- cut --- However, we have found that it is in fact possible to trigger the handling of CFFv1 in AFDKO, by appending the old style "CFF " table to a variable font which already includes a "CFF2" one. This causes DirectWrite to correctly load the variable font, but AFDKO still finds "CFF " first and passes it for further parsing, thanks to the following logic in srcOpen() (afdko/c/public/lib/source/cffread/cffread.c): --- cut --- 561 /* OTF; use CFF table offset */ 562 sfrTable *table = 563 sfrGetTableByTag(h->ctx.sfr, CTL_TAG('C', 'F', 'F', ' ')); 564 if (table == NULL) { 565 table = sfrGetTableByTag(h->ctx.sfr, CTL_TAG('C', 'F', 'F', '2')); 566 } 567 if (table == NULL) 568 fatal(h, cfrErrNoCFF); 569 origin = table->offset; --- cut --- -----=====[ Crash logs ]=====----- A 64-bit build of "tx" compiled with AddressSanitizer, started with ./tx -cff poc.otf prints out the following report: --- cut --- ================================================================= ==236657==ERROR: AddressSanitizer: heap-buffer-overflow on address 0x62a00000b228 at pc 0x0000005563be bp 0x7ffe3c238d10 sp 0x7ffe3c238d08 WRITE of size 2 at 0x62a00000b228 thread T0 #0 0x5563bd in addID afdko/c/public/lib/source/cffread/cffread.c:1843:19 #1 0x53f71c in readCharset afdko/c/public/lib/source/cffread/cffread.c:2187:29 #2 0x5299c7 in cfrBegFont afdko/c/public/lib/source/cffread/cffread.c:2789:9 #3 0x50928d in cfrReadFont afdko/c/tx/source/tx.c:137:9 #4 0x508cc3 in doFile afdko/c/tx/source/tx.c:429:17 #5 0x506b2e in doSingleFileSet afdko/c/tx/source/tx.c:488:5 #6 0x4fc91e in parseArgs afdko/c/tx/source/tx.c:558:17 #7 0x4f9470 in main afdko/c/tx/source/tx.c:1631:9 #8 0x7f7bf34352b0 in __libc_start_main #9 0x41e5b9 in _start 0x62a00000b228 is located 40 bytes to the right of 20480-byte region [0x62a000006200,0x62a00000b200) allocated by thread T0 here: #0 0x4c63f3 in __interceptor_malloc #1 0x6c9ac2 in mem_manage afdko/c/public/lib/source/tx_shared/tx_shared.c:73:20 #2 0x5474a4 in dna_manage afdko/c/public/lib/source/cffread/cffread.c:271:17 #3 0x7de64e in dnaGrow afdko/c/public/lib/source/dynarr/dynarr.c:86:23 #4 0x7dec75 in dnaSetCnt afdko/c/public/lib/source/dynarr/dynarr.c:119:13 #5 0x53e6fa in readCharStringsINDEX afdko/c/public/lib/source/cffread/cffread.c:1791:5 #6 0x5295be in cfrBegFont afdko/c/public/lib/source/cffread/cffread.c:2769:9 #7 0x50928d in cfrReadFont afdko/c/tx/source/tx.c:137:9 #8 0x508cc3 in doFile afdko/c/tx/source/tx.c:429:17 #9 0x506b2e in doSingleFileSet afdko/c/tx/source/tx.c:488:5 #10 0x4fc91e in parseArgs afdko/c/tx/source/tx.c:558:17 #11 0x4f9470 in main afdko/c/tx/source/tx.c:1631:9 #12 0x7f7bf34352b0 in __libc_start_main SUMMARY: AddressSanitizer: heap-buffer-overflow afdko/c/public/lib/source/cffread/cffread.c:1843:19 in addID Shadow bytes around the buggy address: 0x0c547fff95f0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0x0c547fff9600: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0x0c547fff9610: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0x0c547fff9620: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0x0c547fff9630: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 =>0x0c547fff9640: fa fa fa fa fa[fa]fa fa fa fa fa fa fa fa fa fa 0x0c547fff9650: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa 0x0c547fff9660: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa 0x0c547fff9670: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa 0x0c547fff9680: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa 0x0c547fff9690: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa Shadow byte legend (one shadow byte represents 8 application bytes): Addressable: 00 Partially addressable: 01 02 03 04 05 06 07 Heap left redzone: fa Freed heap region: fd Stack left redzone: f1 Stack mid redzone: f2 Stack right redzone: f3 Stack after return: f5 Stack use after scope: f8 Global redzone: f9 Global init order: f6 Poisoned by user: f7 Container overflow: fc Array cookie: ac Intra object redzone: bb ASan internal: fe Left alloca redzone: ca Right alloca redzone: cb Shadow gap: cc ==236657==ABORTING --- cut --- A non-instrumented version of "tx" crashes with a SIGSEGV when it reaches an unmapped memory area: --- cut --- Program received signal SIGSEGV, Segmentation fault. 0x0000000000417d1e in addID (h=0x7103a0, gid=2293, id=18997) at ../../../../../source/cffread/cffread.c:1843 1843 info->cid = id; (gdb) print info $1 = (abfGlyphInfo *) 0x743000 (gdb) print &h->glyphs.array[gid] $2 = (abfGlyphInfo *) 0x743000 (gdb) print gid $3 = 2293 (gdb) x/10gx 0x743000 0x743000: Cannot access memory at address 0x743000 (gdb) bt #0 0x0000000000417d1e in addID (h=0x7103a0, gid=2293, id=18997) at ../../../../../source/cffread/cffread.c:1843 #1 0x0000000000412a57 in readCharset (h=0x7103a0) at ../../../../../source/cffread/cffread.c:2187 #2 0x000000000040dd64 in cfrBegFont (h=0x7103a0, flags=4, origin=0, ttcIndex=0, top=0x6f6048, UDV=0x0) at ../../../../../source/cffread/cffread.c:2789 #3 0x0000000000405e4e in cfrReadFont (h=0x6f6010, origin=0, ttcIndex=0) at ../../../../source/tx.c:137 #4 0x0000000000405c9e in doFile (h=0x6f6010, srcname=0x7fffffffdf4c "poc.otf") at ../../../../source/tx.c:429 #5 0x000000000040532e in doSingleFileSet (h=0x6f6010, srcname=0x7fffffffdf4c "poc.otf") at ../../../../source/tx.c:488 #6 0x0000000000402f59 in parseArgs (h=0x6f6010, argc=2, argv=0x7fffffffdc50) at ../../../../source/tx.c:558 #7 0x0000000000401df2 in main (argc=2, argv=0x7fffffffdc50) at ../../../../source/tx.c:1631 (gdb) --- cut --- A similar Microsoft Edge renderer process crash is also shown below: --- cut --- (4d58.50bc): Access violation - code c0000005 (first chance) First chance exceptions are reported before any exception handling. This exception may be expected and handled. DWrite!addID+0x33: 00007ffb`29e6864b 66895cfe28 mov word ptr [rsi+rdi*8+28h],bx ds:000001ea`f5fee000=???? 0:038> ? rsi Evaluate expression: 2108661076032 = 000001ea`f5fe9040 0:038> ? rdi Evaluate expression: 2547 = 00000000`000009f3 0:038> db rsi+rdi*8+28 000001ea`f5fee000 ?? ?? ?? ?? ?? ?? ?? ??-?? ?? ?? ?? ?? ?? ?? ?? ???????????????? 000001ea`f5fee010 ?? ?? ?? ?? ?? ?? ?? ??-?? ?? ?? ?? ?? ?? ?? ?? ???????????????? 000001ea`f5fee020 ?? ?? ?? ?? ?? ?? ?? ??-?? ?? ?? ?? ?? ?? ?? ?? ???????????????? 000001ea`f5fee030 ?? ?? ?? ?? ?? ?? ?? ??-?? ?? ?? ?? ?? ?? ?? ?? ???????????????? 000001ea`f5fee040 ?? ?? ?? ?? ?? ?? ?? ??-?? ?? ?? ?? ?? ?? ?? ?? ???????????????? 000001ea`f5fee050 ?? ?? ?? ?? ?? ?? ?? ??-?? ?? ?? ?? ?? ?? ?? ?? ???????????????? 000001ea`f5fee060 ?? ?? ?? ?? ?? ?? ?? ??-?? ?? ?? ?? ?? ?? ?? ?? ???????????????? 000001ea`f5fee070 ?? ?? ?? ?? ?? ?? ?? ??-?? ?? ?? ?? ?? ?? ?? ?? ???????????????? 0:038> k # Child-SP RetAddr Call Site 00 00000047`0fcfae10 00007ffb`29e6a262 DWrite!addID+0x33 01 00000047`0fcfae40 00007ffb`29e6de84 DWrite!readCharset+0x10a 02 00000047`0fcfae70 00007ffb`29e621e7 DWrite!cfrBegFont+0x5d8 03 00000047`0fcfb700 00007ffb`29df157a DWrite!AdobeCFF2Snapshot+0x10f 04 00000047`0fcfbc00 00007ffb`29df0729 DWrite!FontInstancer::InstanceCffTable+0x212 05 00000047`0fcfbde0 00007ffb`29df039a DWrite!FontInstancer::CreateInstanceInternal+0x249 06 00000047`0fcfc000 00007ffb`29dd5a4e DWrite!FontInstancer::CreateInstance+0x192 07 00000047`0fcfc360 00007ffb`34eb61ab DWrite!DWriteFontFace::CreateInstancedStream+0x9e 08 00000047`0fcfc3f0 00007ffb`34ea9148 d2d1!dxc::TextConvertor::InstanceFontResources+0x19f 09 00000047`0fcfc510 00007ffb`0fb750f4 d2d1!dxc::CXpsPrintControl::Close+0xc8 0a 00000047`0fcfc560 00007ffb`0fb4fcb0 edgehtml!CDXPrintControl::Close+0x44 0b 00000047`0fcfc5b0 00007ffb`0fb547ad edgehtml!CTemplatePrinter::EndPrintD2D+0x5c 0c 00000047`0fcfc5e0 00007ffb`0fa2b515 edgehtml!CPrintManagerTemplatePrinter::endPrint+0x2d 0d 00000047`0fcfc610 00007ffb`0f689175 edgehtml!CFastDOM::CMSPrintManagerTemplatePrinter::Trampoline_endPrint+0x45 0e 00000047`0fcfc650 00007ffb`0eb568f1 edgehtml!CFastDOM::CMSPrintManagerTemplatePrinter::Profiler_endPrint+0x25 [...] --- cut --- -----=====[ References ]=====----- [1] https://blog.typekit.com/2014/09/19/new-from-adobe-type-open-sourced-font-development-tools/ [2] https://github.com/adobe-type-tools/afdko [3] https://docs.microsoft.com/en-us/windows/desktop/directwrite/direct-write-portal [4] https://medium.com/variable-fonts/https-medium-com-tiro-introducing-opentype-variable-fonts-12ba6cd2369 Proof of Concept: https://gitlab.com/exploit-database/exploitdb-bin-sploits/-/raw/main/bin-sploits/47095.zip |