diff options
Diffstat (limited to 'meta/recipes-devtools/binutils')
8 files changed, 2432 insertions, 546 deletions
diff --git a/meta/recipes-devtools/binutils/binutils-2.36.inc b/meta/recipes-devtools/binutils/binutils-2.36.inc index 9d770db5a8..fa28358c2d 100644 --- a/meta/recipes-devtools/binutils/binutils-2.36.inc +++ b/meta/recipes-devtools/binutils/binutils-2.36.inc @@ -24,7 +24,7 @@ BRANCH ?= "binutils-2_36-branch" UPSTREAM_CHECK_GITTAGREGEX = "binutils-(?P<pver>\d+_(\d_?)*)" -SRCREV ?= "7651a4871c225925ffdfda0a8c91a6ed370cd9a1" +SRCREV ?= "a281816c8aeb12619d34eec8959a43dfa5c6b4ec" BINUTILS_GIT_URI ?= "git://sourceware.org/git/binutils-gdb.git;branch=${BRANCH};protocol=git" SRC_URI = "\ ${BINUTILS_GIT_URI} \ @@ -41,8 +41,9 @@ SRC_URI = "\ file://0014-Fix-rpath-in-libtool-when-sysroot-is-enabled.patch \ file://0015-sync-with-OE-libtool-changes.patch \ file://0016-Check-for-clang-before-checking-gcc-version.patch \ - file://0001-CVE-2021-20197.patch \ - file://0002-CVE-2021-20197.patch \ - file://0003-CVE-2021-20197.patch \ + file://0017-CVE-2021-3530.patch \ + file://0018-CVE-2021-3530.patch \ + file://0001-CVE-2021-42574.patch \ + file://0001-CVE-2021-45078.patch \ " S = "${WORKDIR}/git" diff --git a/meta/recipes-devtools/binutils/binutils/0001-CVE-2021-20197.patch b/meta/recipes-devtools/binutils/binutils/0001-CVE-2021-20197.patch deleted file mode 100644 index 2b4eaba26d..0000000000 --- a/meta/recipes-devtools/binutils/binutils/0001-CVE-2021-20197.patch +++ /dev/null @@ -1,201 +0,0 @@ -From 8e03235147a9e774d3ba084e93c2da1aa94d1cec Mon Sep 17 00:00:00 2001 -From: Siddhesh Poyarekar <siddhesh@gotplt.org> -Date: Mon, 22 Feb 2021 20:45:50 +0530 -Subject: [PATCH] binutils: Avoid renaming over existing files - -Renaming over existing files needs additional care to restore -permissions and ownership, which may not always succeed. -Additionally, other properties of the file such as extended attributes -may be lost, making the operation flaky. - -For predictable results, resort to rename() only if the file does not -exist, otherwise copy the file contents into the existing file. This -ensures that no additional tricks are needed to retain file -properties. - -This also allows dropping of the redundant set_times on the tmpfile in -objcopy/strip since now we no longer rename over existing files. - -binutils/ - - * ar.c (write_archive): Adjust call to SMART_RENAME. - * arsup.c (ar_save): Likewise. - * objcopy (strip_main): Don't set times on temporary file and - adjust call to SMART_RENAME. - (copy_main): Likewise. - * rename.c [!S_ISLNK]: Remove definitions. - (try_preserve_permissions): Remove function. - (smart_rename): Replace PRESERVE_DATES argument with - TARGET_STAT. Use rename system call only if TO does not exist. - * bucomm.h (smart_rename): Adjust declaration. - -(cherry picked from commit 3685de750e6a091663a0abe42528cad29e960e35) - -Upstream-Status: Backport [https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;h=8e03235147a9e774d3ba084e93c2da1aa94d1cec] -CVE: CVE-2021-20197 -Signed-off-by: Vinay Kumar <vinay.m.engg@gmail.com> ---- - binutils/ar.c | 2 +- - binutils/arsup.c | 2 +- - binutils/bucomm.h | 3 ++- - binutils/objcopy.c | 8 ++----- - binutils/rename.c | 55 +++++++++------------------------------------- - 6 files changed, 29 insertions(+), 54 deletions(-) - -diff --git a/binutils/ar.c b/binutils/ar.c -index 45a34e3a6cf..3a91708b51c 100644 ---- a/binutils/ar.c -+++ b/binutils/ar.c -@@ -1308,7 +1308,7 @@ write_archive (bfd *iarch) - /* We don't care if this fails; we might be creating the archive. */ - bfd_close (iarch); - -- if (smart_rename (new_name, old_name, 0) != 0) -+ if (smart_rename (new_name, old_name, NULL) != 0) - xexit (1); - free (old_name); - free (new_name); -diff --git a/binutils/arsup.c b/binutils/arsup.c -index 5403a0c5d74..0a1f63f6456 100644 ---- a/binutils/arsup.c -+++ b/binutils/arsup.c -@@ -351,7 +351,7 @@ ar_save (void) - - bfd_close (obfd); - -- smart_rename (ofilename, real_name, 0); -+ smart_rename (ofilename, real_name, NULL); - obfd = 0; - free (ofilename); - } -diff --git a/binutils/bucomm.h b/binutils/bucomm.h -index 91f6a5b228f..aa7e33d8cd1 100644 ---- a/binutils/bucomm.h -+++ b/binutils/bucomm.h -@@ -71,7 +71,8 @@ extern void print_version (const char *); - /* In rename.c. */ - extern void set_times (const char *, const struct stat *); - --extern int smart_rename (const char *, const char *, int); -+extern int smart_rename (const char *, const char *, struct stat *); -+ - - /* In libiberty. */ - void *xmalloc (size_t); -diff --git a/binutils/objcopy.c b/binutils/objcopy.c -index eab3b6db585..07a872b5a80 100644 ---- a/binutils/objcopy.c -+++ b/binutils/objcopy.c -@@ -4861,12 +4861,10 @@ strip_main (int argc, char *argv[]) - output_target, NULL); - if (status == 0) - { -- if (preserve_dates) -- set_times (tmpname, &statbuf); - if (output_file != tmpname) - status = (smart_rename (tmpname, - output_file ? output_file : argv[i], -- preserve_dates) != 0); -+ preserve_dates ? &statbuf : NULL) != 0); - if (status == 0) - status = hold_status; - } -@@ -5931,11 +5929,9 @@ copy_main (int argc, char *argv[]) - output_target, input_arch); - if (status == 0) - { -- if (preserve_dates) -- set_times (tmpname, &statbuf); - if (tmpname != output_filename) - status = (smart_rename (tmpname, input_filename, -- preserve_dates) != 0); -+ preserve_dates ? &statbuf : NULL) != 0); - } - else - unlink_if_ordinary (tmpname); -diff --git a/binutils/rename.c b/binutils/rename.c -index 65ad5bf52c4..f471b45fd3f 100644 ---- a/binutils/rename.c -+++ b/binutils/rename.c -@@ -122,20 +122,13 @@ set_times (const char *destination, const struct stat *statbuf) - non_fatal (_("%s: cannot set time: %s"), destination, strerror (errno)); - } - --#ifndef S_ISLNK --#ifdef S_IFLNK --#define S_ISLNK(m) (((m) & S_IFMT) == S_IFLNK) --#else --#define S_ISLNK(m) 0 --#define lstat stat --#endif --#endif -- --/* Rename FROM to TO, copying if TO is a link. -- Return 0 if ok, -1 if error. */ -+/* Rename FROM to TO, copying if TO exists. TARGET_STAT has the file status -+ that, if non-NULL, is used to fix up timestamps after rename. Return 0 if -+ ok, -1 if error. */ - - int --smart_rename (const char *from, const char *to, int preserve_dates ATTRIBUTE_UNUSED) -+smart_rename (const char *from, const char *to, -+ struct stat *target_stat ATTRIBUTE_UNUSED) - { - bfd_boolean exists; - struct stat s; -@@ -158,38 +151,10 @@ smart_rename (const char *from, const char *to, int preserve_dates ATTRIBUTE_UNU - unlink (from); - } - #else -- /* Use rename only if TO is not a symbolic link and has -- only one hard link, and we have permission to write to it. */ -- if (! exists -- || (!S_ISLNK (s.st_mode) -- && S_ISREG (s.st_mode) -- && (s.st_mode & S_IWUSR) -- && s.st_nlink == 1) -- ) -+ /* Avoid a full copy and use rename if TO does not exist. */ -+ if (!exists) - { -- ret = rename (from, to); -- if (ret == 0) -- { -- if (exists) -- { -- /* Try to preserve the permission bits and ownership of -- TO. First get the mode right except for the setuid -- bit. Then change the ownership. Then fix the setuid -- bit. We do the chmod before the chown because if the -- chown succeeds, and we are a normal user, we won't be -- able to do the chmod afterward. We don't bother to -- fix the setuid bit first because that might introduce -- a fleeting security problem, and because the chown -- will clear the setuid bit anyhow. We only fix the -- setuid bit if the chown succeeds, because we don't -- want to introduce an unexpected setuid file owned by -- the user running objcopy. */ -- chmod (to, s.st_mode & 0777); -- if (chown (to, s.st_uid, s.st_gid) >= 0) -- chmod (to, s.st_mode & 07777); -- } -- } -- else -+ if ((ret = rename (from, to)) != 0) - { - /* We have to clean up here. */ - non_fatal (_("unable to rename '%s'; reason: %s"), to, strerror (errno)); -@@ -202,8 +167,8 @@ smart_rename (const char *from, const char *to, int preserve_dates ATTRIBUTE_UNU - if (ret != 0) - non_fatal (_("unable to copy file '%s'; reason: %s"), to, strerror (errno)); - -- if (preserve_dates) -- set_times (to, &s); -+ if (target_stat != NULL) -+ set_times (to, target_stat); - unlink (from); - } - #endif /* _WIN32 && !__CYGWIN32__ */ --- -2.31.1 - diff --git a/meta/recipes-devtools/binutils/binutils/0001-CVE-2021-42574.patch b/meta/recipes-devtools/binutils/binutils/0001-CVE-2021-42574.patch new file mode 100644 index 0000000000..6adc438b5c --- /dev/null +++ b/meta/recipes-devtools/binutils/binutils/0001-CVE-2021-42574.patch @@ -0,0 +1,2006 @@ +From b3aa80b45c4f46029efeb204bb9f2d2c4278a0e5 Mon Sep 17 00:00:00 2001 +From: Nick Clifton <nickc@redhat.com> +Date: Tue, 9 Nov 2021 13:25:42 +0000 +Subject: [PATCH] Add --unicode option to control how unicode characters are + handled by display tools. + + * nm.c: Add --unicode option to control how unicode characters are + handled. + * objdump.c: Likewise. + * readelf.c: Likewise. + * strings.c: Likewise. + * binutils.texi: Document the new feature. + * NEWS: Document the new feature. + * testsuite/binutils-all/unicode.exp: New file. + * testsuite/binutils-all/nm.hex.unicode + * testsuite/binutils-all/strings.escape.unicode + * testsuite/binutils-all/objdump.highlight.unicode + * testsuite/binutils-all/readelf.invalid.unicode + +CVE: CVE-2021-42574 +Upstream-Status: Backport [https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=b3aa80b45c4f46029efeb204bb9f2d2c4278a0e5] + +Signed-off-by: pgowda <pgowda.cve@gmail.com> +--- + binutils/ChangeLog | 15 + + binutils/NEWS | 9 + + binutils/doc/binutils.texi | 78 ++++ + binutils/nm.c | 228 ++++++++++- + binutils/objdump.c | 235 ++++++++++-- + binutils/readelf.c | 190 +++++++++- + binutils/strings.c | 757 ++++++++++++++++++++++++++++++++++--- + 7 files changed, 1409 insertions(+), 103 deletions(-) + +diff --git a/binutils/ChangeLog b/binutils/ChangeLog +--- a/binutils/ChangeLog 2021-12-23 03:23:38.425843662 -0800 ++++ b/binutils/ChangeLog 2021-12-23 03:30:31.094968942 -0800 +@@ -1,3 +1,18 @@ ++2021-11-09 Nick Clifton <nickc@redhat.com> ++ ++ * nm.c: Add --unicode option to control how unicode characters are ++ handled. ++ * objdump.c: Likewise. ++ * readelf.c: Likewise. ++ * strings.c: Likewise. ++ * binutils.texi: Document the new feature. ++ * NEWS: Document the new feature. ++ * testsuite/binutils-all/unicode.exp: New file. ++ * testsuite/binutils-all/nm.hex.unicode ++ * testsuite/binutils-all/strings.escape.unicode ++ * testsuite/binutils-all/objdump.highlight.unicode ++ * testsuite/binutils-all/readelf.invalid.unicode ++ + 2021-04-15 Alan Modra <amodra@gmail.com> + + PR 27456 +diff --git a/binutils/doc/binutils.texi b/binutils/doc/binutils.texi +--- a/binutils/doc/binutils.texi 2021-12-23 03:23:38.441843395 -0800 ++++ b/binutils/doc/binutils.texi 2021-12-23 03:30:31.094968942 -0800 +@@ -805,6 +805,7 @@ nm [@option{-A}|@option{-o}|@option{--pr + [@option{-s}|@option{--print-armap}] [@option{-t} @var{radix}|@option{--radix=}@var{radix}] + [@option{-u}|@option{--undefined-only}] [@option{-V}|@option{--version}] + [@option{-X 32_64}] [@option{--defined-only}] [@option{--no-demangle}] ++ [@option{-U} @var{method}] [@option{--unicode=}@var{method}] + [@option{--plugin} @var{name}] + [@option{--no-recurse-limit}|@option{--recurse-limit}]] + [@option{--size-sort}] [@option{--special-syms}] +@@ -1114,6 +1115,21 @@ Use @var{radix} as the radix for printin + @cindex undefined symbols + Display only undefined symbols (those external to each object file). + ++@item -U @var{[d|i|l|e|x|h]} ++@itemx --unicode=@var{[default|invalid|locale|escape|hex|highlight]} ++Controls the display of UTF-8 encoded mulibyte characters in strings. ++The default (@option{--unicode=default}) is to give them no special ++treatment. The @option{--unicode=locale} option displays the sequence ++in the current locale, which may or may not support them. The options ++@option{--unicode=hex} and @option{--unicode=invalid} display them as ++hex byte sequences enclosed by either angle brackets or curly braces. ++ ++The @option{--unicode=escape} option displays them as escape sequences ++(@var{\uxxxx}) and the @option{--unicode=highlight} option displays ++them as escape sequences highlighted in red (if supported by the ++output device). The colouring is intended to draw attention to the ++presence of unicode sequences where they might not be expected. ++ + @item -V + @itemx --version + Show the version number of @command{nm} and exit. +@@ -2210,6 +2226,7 @@ objdump [@option{-a}|@option{--archive-h + [@option{--prefix-strip=}@var{level}] + [@option{--insn-width=}@var{width}] + [@option{--visualize-jumps[=color|=extended-color|=off]} ++ [@option{-U} @var{method}] [@option{--unicode=}@var{method}] + [@option{-V}|@option{--version}] + [@option{-H}|@option{--help}] + @var{objfile}@dots{} +@@ -2877,6 +2894,21 @@ When displaying symbols include those wh + special in some way and which would not normally be of interest to the + user. + ++@item -U @var{[d|i|l|e|x|h]} ++@itemx --unicode=@var{[default|invalid|locale|escape|hex|highlight]} ++Controls the display of UTF-8 encoded mulibyte characters in strings. ++The default (@option{--unicode=default}) is to give them no special ++treatment. The @option{--unicode=locale} option displays the sequence ++in the current locale, which may or may not support them. The options ++@option{--unicode=hex} and @option{--unicode=invalid} display them as ++hex byte sequences enclosed by either angle brackets or curly braces. ++ ++The @option{--unicode=escape} option displays them as escape sequences ++(@var{\uxxxx}) and the @option{--unicode=highlight} option displays ++them as escape sequences highlighted in red (if supported by the ++output device). The colouring is intended to draw attention to the ++presence of unicode sequences where they might not be expected. ++ + @item -V + @itemx --version + Print the version number of @command{objdump} and exit. +@@ -3153,6 +3185,7 @@ strings [@option{-afovV}] [@option{-}@va + [@option{-n} @var{min-len}] [@option{--bytes=}@var{min-len}] + [@option{-t} @var{radix}] [@option{--radix=}@var{radix}] + [@option{-e} @var{encoding}] [@option{--encoding=}@var{encoding}] ++ [@option{-U} @var{method}] [@option{--unicode=}@var{method}] + [@option{-}] [@option{--all}] [@option{--print-file-name}] + [@option{-T} @var{bfdname}] [@option{--target=}@var{bfdname}] + [@option{-w}] [@option{--include-all-whitespace}] +@@ -3244,6 +3277,28 @@ single-8-bit-byte characters, @samp{b} = + littleendian. Useful for finding wide character strings. (@samp{l} + and @samp{b} apply to, for example, Unicode UTF-16/UCS-2 encodings). + ++@item -U @var{[d|i|l|e|x|h]} ++@itemx --unicode=@var{[default|invalid|locale|escape|hex|highlight]} ++Controls the display of UTF-8 encoded mulibyte characters in strings. ++The default (@option{--unicode=default}) is to give them no special ++treatment, and instead rely upon the setting of the ++@option{--encoding} option. The other values for this option ++automatically enable @option{--encoding=S}. ++ ++The @option{--unicode=invalid} option treats them as non-graphic ++characters and hence not part of a valid string. All the remaining ++options treat them as valid string characters. ++ ++The @option{--unicode=locale} option displays them in the current ++locale, which may or may not support UTF-8 encoding. The ++@option{--unicode=hex} option displays them as hex byte sequences ++enclosed between @var{<>} characters. The @option{--unicode=escape} ++option displays them as escape sequences (@var{\uxxxx}) and the ++@option{--unicode=highlight} option displays them as escape sequences ++highlighted in red (if supported by the output device). The colouring ++is intended to draw attention to the presence of unicode sequences ++where they might not be expected. ++ + @item -T @var{bfdname} + @itemx --target=@var{bfdname} + @cindex object code format +@@ -4740,6 +4795,7 @@ readelf [@option{-a}|@option{--all}] + [@option{--dyn-syms}|@option{--lto-syms}] + [@option{--demangle@var{=style}}|@option{--no-demangle}] + [@option{--recurse-limit}|@option{--no-recurse-limit}] ++ [@option{-U} @var{method}|@option{--unicode=}@var{method}] + [@option{-n}|@option{--notes}] + [@option{-r}|@option{--relocs}] + [@option{-u}|@option{--unwind}] +@@ -4887,6 +4943,28 @@ necessary in order to demangle truly com + that if the recursion limit is disabled then stack exhaustion is + possible and any bug reports about such an event will be rejected. + ++@item -U @var{[d|i|l|e|x|h]} ++@itemx --unicode=[default|invalid|locale|escape|hex|highlight] ++Controls the display of non-ASCII characters in identifier names. ++The default (@option{--unicode=locale} or @option{--unicode=default}) is ++to treat them as multibyte characters and display them in the current ++locale. All other versions of this option treat the bytes as UTF-8 ++encoded values and attempt to interpret them. If they cannot be ++interpreted or if the @option{--unicode=invalid} option is used then ++they are displayed as a sequence of hex bytes, encloses in curly ++parethesis characters. ++ ++Using the @option{--unicode=escape} option will display the characters ++as as unicode escape sequences (@var{\uxxxx}). Using the ++@option{--unicode=hex} will display the characters as hex byte ++sequences enclosed between angle brackets. ++ ++Using the @option{--unicode=highlight} will display the characters as ++unicode escape sequences but it will also highlighted them in red, ++assuming that colouring is supported by the output device. The ++colouring is intended to draw attention to the presence of unicode ++sequences when they might not be expected. ++ + @item -e + @itemx --headers + Display all the headers in the file. Equivalent to @option{-h -l -S}. +diff --git a/binutils/NEWS b/binutils/NEWS +--- a/binutils/NEWS 2021-12-23 03:23:38.433843528 -0800 ++++ b/binutils/NEWS 2021-12-23 03:30:31.094968942 -0800 +@@ -1,5 +1,14 @@ + -*- text -*- + ++* Tools which display symbols or strings (readelf, strings, nm, objdump) ++ have a new command line option which controls how unicode characters are ++ handled. By default they are treated as normal for the tool. Using ++ --unicode=locale will display them according to the current locale. ++ Using --unicode=hex will display them as hex byte values, whilst ++ --unicode=escape will display them as escape sequences. In addition ++ using --unicode=highlight will display them as unicode escape sequences ++ highlighted in red (if supported by the output device). ++ + Changes in 2.36: + + * Update elfedit and readelf with LAM_U48 and LAM_U57 support. +diff --git a/binutils/nm.c b/binutils/nm.c +--- a/binutils/nm.c 2021-12-23 03:23:38.441843395 -0800 ++++ b/binutils/nm.c 2021-12-23 03:30:31.098968875 -0800 +@@ -38,6 +38,11 @@ + #include "bucomm.h" + #include "plugin-api.h" + #include "plugin.h" ++#include "safe-ctype.h" ++ ++#ifndef streq ++#define streq(a,b) (strcmp ((a),(b)) == 0) ++#endif + + /* When sorting by size, we use this structure to hold the size and a + pointer to the minisymbol. */ +@@ -192,6 +197,18 @@ static const char *plugin_target = NULL; + static bfd *lineno_cache_bfd; + static bfd *lineno_cache_rel_bfd; + ++typedef enum unicode_display_type ++{ ++ unicode_default = 0, ++ unicode_locale, ++ unicode_escape, ++ unicode_hex, ++ unicode_highlight, ++ unicode_invalid ++} unicode_display_type; ++ ++static unicode_display_type unicode_display = unicode_default; ++ + enum long_option_values + { + OPTION_TARGET = 200, +@@ -234,6 +251,7 @@ static struct option long_options[] = + {"target", required_argument, 0, OPTION_TARGET}, + {"defined-only", no_argument, &defined_only, 1}, + {"undefined-only", no_argument, &undefined_only, 1}, ++ {"unicode", required_argument, NULL, 'U'}, + {"version", no_argument, &show_version, 1}, + {"with-symbol-versions", no_argument, NULL, + OPTION_WITH_SYMBOL_VERSIONS}, +@@ -285,6 +303,8 @@ usage (FILE *stream, int status) + -t, --radix=RADIX Use RADIX for printing symbol values\n\ + --target=BFDNAME Specify the target object format as BFDNAME\n\ + -u, --undefined-only Display only undefined symbols\n\ ++ -U {d|s|i|x|e|h} Specify how to treat UTF-8 encoded unicode characters\n\ ++ --unicode={default|show|invalid|hex|escape|highlight}\n\ + --with-symbol-versions Display version strings after symbol names\n\ + -X 32_64 (ignored)\n\ + @FILE Read options from FILE\n\ +@@ -399,6 +419,188 @@ get_coff_symbol_type (const struct inter + fatal ("%s", xstrerror (errno)); + return bufp; + } ++ ++/* Convert a potential UTF-8 encoded sequence in IN into characters in OUT. ++ The conversion format is controlled by the unicode_display variable. ++ Returns the number of characters added to OUT. ++ Returns the number of bytes consumed from IN in CONSUMED. ++ Always consumes at least one byte and displays at least one character. */ ++ ++static unsigned int ++display_utf8 (const unsigned char * in, char * out, unsigned int * consumed) ++{ ++ char * orig_out = out; ++ unsigned int nchars = 0; ++ unsigned int j; ++ ++ if (unicode_display == unicode_default) ++ goto invalid; ++ ++ if (in[0] < 0xc0) ++ goto invalid; ++ ++ if ((in[1] & 0xc0) != 0x80) ++ goto invalid; ++ ++ if ((in[0] & 0x20) == 0) ++ { ++ nchars = 2; ++ goto valid; ++ } ++ ++ if ((in[2] & 0xc0) != 0x80) ++ goto invalid; ++ ++ if ((in[0] & 0x10) == 0) ++ { ++ nchars = 3; ++ goto valid; ++ } ++ ++ if ((in[3] & 0xc0) != 0x80) ++ goto invalid; ++ ++ nchars = 4; ++ ++ valid: ++ switch (unicode_display) ++ { ++ case unicode_locale: ++ /* Copy the bytes into the output buffer as is. */ ++ memcpy (out, in, nchars); ++ out += nchars; ++ break; ++ ++ case unicode_invalid: ++ case unicode_hex: ++ out += sprintf (out, "%c", unicode_display == unicode_hex ? '<' : '{'); ++ out += sprintf (out, "0x"); ++ for (j = 0; j < nchars; j++) ++ out += sprintf (out, "%02x", in [j]); ++ out += sprintf (out, "%c", unicode_display == unicode_hex ? '>' : '}'); ++ break; ++ ++ case unicode_highlight: ++ if (isatty (1)) ++ out += sprintf (out, "\x1B[31;47m"); /* Red. */ ++ /* Fall through. */ ++ case unicode_escape: ++ switch (nchars) ++ { ++ case 2: ++ out += sprintf (out, "\\u%02x%02x", ++ ((in[0] & 0x1c) >> 2), ++ ((in[0] & 0x03) << 6) | (in[1] & 0x3f)); ++ break; ++ ++ case 3: ++ out += sprintf (out, "\\u%02x%02x", ++ ((in[0] & 0x0f) << 4) | ((in[1] & 0x3c) >> 2), ++ ((in[1] & 0x03) << 6) | ((in[2] & 0x3f))); ++ break; ++ ++ case 4: ++ out += sprintf (out, "\\u%02x%02x%02x", ++ ((in[0] & 0x07) << 6) | ((in[1] & 0x3c) >> 2), ++ ((in[1] & 0x03) << 6) | ((in[2] & 0x3c) >> 2), ++ ((in[2] & 0x03) << 6) | ((in[3] & 0x3f))); ++ break; ++ default: ++ /* URG. */ ++ break; ++ } ++ ++ if (unicode_display == unicode_highlight && isatty (1)) ++ out += sprintf (out, "\033[0m"); /* Default colour. */ ++ break; ++ ++ default: ++ /* URG */ ++ break; ++ } ++ ++ * consumed = nchars; ++ return out - orig_out; ++ ++ invalid: ++ /* Not a valid UTF-8 sequence. */ ++ *out = *in; ++ * consumed = 1; ++ return 1; ++} ++ ++/* Convert any UTF-8 encoded characters in NAME into the form specified by ++ unicode_display. Also converts control characters. Returns a static ++ buffer if conversion was necessary. ++ Code stolen from objdump.c:sanitize_string(). */ ++ ++static const char * ++convert_utf8 (const char * in) ++{ ++ static char * buffer = NULL; ++ static size_t buffer_len = 0; ++ const char * original = in; ++ char * out; ++ ++ /* Paranoia. */ ++ if (in == NULL) ++ return ""; ++ ++ /* See if any conversion is necessary. ++ In the majority of cases it will not be needed. */ ++ do ++ { ++ unsigned char c = *in++; ++ ++ if (c == 0) ++ return original; ++ ++ if (ISCNTRL (c)) ++ break; ++ ++ if (unicode_display != unicode_default && c >= 0xc0) ++ break; ++ } ++ while (1); ++ ++ /* Copy the input, translating as needed. */ ++ in = original; ++ if (buffer_len < (strlen (in) * 9)) ++ { ++ free ((void *) buffer); ++ buffer_len = strlen (in) * 9; ++ buffer = xmalloc (buffer_len + 1); ++ } ++ ++ out = buffer; ++ do ++ { ++ unsigned char c = *in++; ++ ++ if (c == 0) ++ break; ++ ++ if (ISCNTRL (c)) ++ { ++ *out++ = '^'; ++ *out++ = c + 0x40; ++ } ++ else if (unicode_display != unicode_default && c >= 0xc0) ++ { ++ unsigned int num_consumed; ++ ++ out += display_utf8 ((const unsigned char *)(in - 1), out, & num_consumed); ++ in += num_consumed - 1; ++ } ++ else ++ *out++ = c; ++ } ++ while (1); ++ ++ *out = 0; ++ return buffer; ++} ++ + + /* Print symbol name NAME, read from ABFD, with printf format FORM, + demangling it if requested. */ +@@ -418,6 +620,11 @@ print_symname (const char *form, struct + name = alloc; + } + ++ if (unicode_display != unicode_default) ++ { ++ name = convert_utf8 (name); ++ } ++ + if (info != NULL && info->elfinfo) + { + const char *version_string; +@@ -1738,7 +1945,7 @@ main (int argc, char **argv) + fatal (_("fatal error: libbfd ABI mismatch")); + set_default_bfd_target (); + +- while ((c = getopt_long (argc, argv, "aABCDef:gHhlnopPrSst:uvVvX:", ++ while ((c = getopt_long (argc, argv, "aABCDef:gHhlnopPrSst:uU:vVvX:", + long_options, (int *) 0)) != EOF) + { + switch (c) +@@ -1828,6 +2035,24 @@ main (int argc, char **argv) + case 'u': + undefined_only = 1; + break; ++ ++ case 'U': ++ if (streq (optarg, "default") || streq (optarg, "d")) ++ unicode_display = unicode_default; ++ else if (streq (optarg, "locale") || streq (optarg, "l")) ++ unicode_display = unicode_locale; ++ else if (streq (optarg, "escape") || streq (optarg, "e")) ++ unicode_display = unicode_escape; ++ else if (streq (optarg, "invalid") || streq (optarg, "i")) ++ unicode_display = unicode_invalid; ++ else if (streq (optarg, "hex") || streq (optarg, "x")) ++ unicode_display = unicode_hex; ++ else if (streq (optarg, "highlight") || streq (optarg, "h")) ++ unicode_display = unicode_highlight; ++ else ++ fatal (_("invalid argument to -U/--unicode: %s"), optarg); ++ break; ++ + case 'V': + show_version = 1; + break; +diff --git a/binutils/objdump.c b/binutils/objdump.c +--- a/binutils/objdump.c 2021-12-23 03:23:38.445843329 -0800 ++++ b/binutils/objdump.c 2021-12-23 03:30:31.098968875 -0800 +@@ -205,6 +205,19 @@ static const struct objdump_private_desc + + /* The list of detected jumps inside a function. */ + static struct jump_info *detected_jumps = NULL; ++ ++typedef enum unicode_display_type ++{ ++ unicode_default = 0, ++ unicode_locale, ++ unicode_escape, ++ unicode_hex, ++ unicode_highlight, ++ unicode_invalid ++} unicode_display_type; ++ ++static unicode_display_type unicode_display = unicode_default; ++ + + static void usage (FILE *, int) ATTRIBUTE_NORETURN; + static void +@@ -274,6 +287,8 @@ usage (FILE *stream, int status) + --recurse-limit Enable a limit on recursion whilst demangling. [Default]\n\ + --no-recurse-limit Disable a limit on recursion whilst demangling\n\ + -w, --wide Format output for more than 80 columns\n\ ++ -U[d|l|i|x|e|h] Controls the display of UTF-8 unicode characters\n\ ++ --unicode=[default|locale|invalid|hex|escape|highlight]\n\ + -z, --disassemble-zeroes Do not skip blocks of zeroes when disassembling\n\ + --start-address=ADDR Only process data whose address is >= ADDR\n\ + --stop-address=ADDR Only process data whose address is < ADDR\n\ +@@ -348,17 +363,23 @@ static struct option long_options[]= + { + {"adjust-vma", required_argument, NULL, OPTION_ADJUST_VMA}, + {"all-headers", no_argument, NULL, 'x'}, +- {"private-headers", no_argument, NULL, 'p'}, +- {"private", required_argument, NULL, 'P'}, + {"architecture", required_argument, NULL, 'm'}, + {"archive-headers", no_argument, NULL, 'a'}, ++#ifdef ENABLE_LIBCTF ++ {"ctf", required_argument, NULL, OPTION_CTF}, ++ {"ctf-parent", required_argument, NULL, OPTION_CTF_PARENT}, ++#endif + {"debugging", no_argument, NULL, 'g'}, + {"debugging-tags", no_argument, NULL, 'e'}, + {"demangle", optional_argument, NULL, 'C'}, + {"disassemble", optional_argument, NULL, 'd'}, + {"disassemble-all", no_argument, NULL, 'D'}, +- {"disassembler-options", required_argument, NULL, 'M'}, + {"disassemble-zeroes", no_argument, NULL, 'z'}, ++ {"disassembler-options", required_argument, NULL, 'M'}, ++ {"dwarf", optional_argument, NULL, OPTION_DWARF}, ++ {"dwarf-check", no_argument, 0, OPTION_DWARF_CHECK}, ++ {"dwarf-depth", required_argument, 0, OPTION_DWARF_DEPTH}, ++ {"dwarf-start", required_argument, 0, OPTION_DWARF_START}, + {"dynamic-reloc", no_argument, NULL, 'R'}, + {"dynamic-syms", no_argument, NULL, 'T'}, + {"endian", required_argument, NULL, OPTION_ENDIAN}, +@@ -368,15 +389,22 @@ static struct option long_options[]= + {"full-contents", no_argument, NULL, 's'}, + {"headers", no_argument, NULL, 'h'}, + {"help", no_argument, NULL, 'H'}, ++ {"include", required_argument, NULL, 'I'}, + {"info", no_argument, NULL, 'i'}, ++ {"inlines", no_argument, 0, OPTION_INLINES}, ++ {"insn-width", required_argument, NULL, OPTION_INSN_WIDTH}, + {"line-numbers", no_argument, NULL, 'l'}, +- {"no-show-raw-insn", no_argument, &show_raw_insn, -1}, + {"no-addresses", no_argument, &no_addresses, 1}, ++ {"no-recurse-limit", no_argument, NULL, OPTION_NO_RECURSE_LIMIT}, ++ {"no-recursion-limit", no_argument, NULL, OPTION_NO_RECURSE_LIMIT}, ++ {"no-show-raw-insn", no_argument, &show_raw_insn, -1}, ++ {"prefix", required_argument, NULL, OPTION_PREFIX}, + {"prefix-addresses", no_argument, &prefix_addresses, 1}, ++ {"prefix-strip", required_argument, NULL, OPTION_PREFIX_STRIP}, ++ {"private", required_argument, NULL, 'P'}, ++ {"private-headers", no_argument, NULL, 'p'}, + {"recurse-limit", no_argument, NULL, OPTION_RECURSE_LIMIT}, + {"recursion-limit", no_argument, NULL, OPTION_RECURSE_LIMIT}, +- {"no-recurse-limit", no_argument, NULL, OPTION_NO_RECURSE_LIMIT}, +- {"no-recursion-limit", no_argument, NULL, OPTION_NO_RECURSE_LIMIT}, + {"reloc", no_argument, NULL, 'r'}, + {"section", required_argument, NULL, 'j'}, + {"section-headers", no_argument, NULL, 'h'}, +@@ -384,28 +412,16 @@ static struct option long_options[]= + {"source", no_argument, NULL, 'S'}, + {"source-comment", optional_argument, NULL, OPTION_SOURCE_COMMENT}, + {"special-syms", no_argument, &dump_special_syms, 1}, +- {"include", required_argument, NULL, 'I'}, +- {"dwarf", optional_argument, NULL, OPTION_DWARF}, +-#ifdef ENABLE_LIBCTF +- {"ctf", required_argument, NULL, OPTION_CTF}, +- {"ctf-parent", required_argument, NULL, OPTION_CTF_PARENT}, +-#endif + {"stabs", no_argument, NULL, 'G'}, + {"start-address", required_argument, NULL, OPTION_START_ADDRESS}, + {"stop-address", required_argument, NULL, OPTION_STOP_ADDRESS}, + {"syms", no_argument, NULL, 't'}, + {"target", required_argument, NULL, 'b'}, ++ {"unicode", required_argument, NULL, 'U'}, + {"version", no_argument, NULL, 'V'}, +- {"wide", no_argument, NULL, 'w'}, +- {"prefix", required_argument, NULL, OPTION_PREFIX}, +- {"prefix-strip", required_argument, NULL, OPTION_PREFIX_STRIP}, +- {"insn-width", required_argument, NULL, OPTION_INSN_WIDTH}, +- {"dwarf-depth", required_argument, 0, OPTION_DWARF_DEPTH}, +- {"dwarf-start", required_argument, 0, OPTION_DWARF_START}, +- {"dwarf-check", no_argument, 0, OPTION_DWARF_CHECK}, +- {"inlines", no_argument, 0, OPTION_INLINES}, + {"visualize-jumps", optional_argument, 0, OPTION_VISUALIZE_JUMPS}, +- {0, no_argument, 0, 0} ++ {"wide", no_argument, NULL, 'w'}, ++ {NULL, no_argument, NULL, 0} + }; + + static void +@@ -415,9 +431,121 @@ nonfatal (const char *msg) + exit_status = 1; + } + ++/* Convert a potential UTF-8 encoded sequence in IN into characters in OUT. ++ The conversion format is controlled by the unicode_display variable. ++ Returns the number of characters added to OUT. ++ Returns the number of bytes consumed from IN in CONSUMED. ++ Always consumes at least one byte and displays at least one character. */ ++ ++static unsigned int ++display_utf8 (const unsigned char * in, char * out, unsigned int * consumed) ++{ ++ char * orig_out = out; ++ unsigned int nchars = 0; ++ unsigned int j; ++ ++ if (unicode_display == unicode_default) ++ goto invalid; ++ ++ if (in[0] < 0xc0) ++ goto invalid; ++ ++ if ((in[1] & 0xc0) != 0x80) ++ goto invalid; ++ ++ if ((in[0] & 0x20) == 0) ++ { ++ nchars = 2; ++ goto valid; ++ } ++ ++ if ((in[2] & 0xc0) != 0x80) ++ goto invalid; ++ ++ if ((in[0] & 0x10) == 0) ++ { ++ nchars = 3; ++ goto valid; ++ } ++ ++ if ((in[3] & 0xc0) != 0x80) ++ goto invalid; ++ ++ nchars = 4; ++ ++ valid: ++ switch (unicode_display) ++ { ++ case unicode_locale: ++ /* Copy the bytes into the output buffer as is. */ ++ memcpy (out, in, nchars); ++ out += nchars; ++ break; ++ ++ case unicode_invalid: ++ case unicode_hex: ++ out += sprintf (out, "%c", unicode_display == unicode_hex ? '<' : '{'); ++ out += sprintf (out, "0x"); ++ for (j = 0; j < nchars; j++) ++ out += sprintf (out, "%02x", in [j]); ++ out += sprintf (out, "%c", unicode_display == unicode_hex ? '>' : '}'); ++ break; ++ ++ case unicode_highlight: ++ if (isatty (1)) ++ out += sprintf (out, "\x1B[31;47m"); /* Red. */ ++ /* Fall through. */ ++ case unicode_escape: ++ switch (nchars) ++ { ++ case 2: ++ out += sprintf (out, "\\u%02x%02x", ++ ((in[0] & 0x1c) >> 2), ++ ((in[0] & 0x03) << 6) | (in[1] & 0x3f)); ++ break; ++ ++ case 3: ++ out += sprintf (out, "\\u%02x%02x", ++ ((in[0] & 0x0f) << 4) | ((in[1] & 0x3c) >> 2), ++ ((in[1] & 0x03) << 6) | ((in[2] & 0x3f))); ++ break; ++ ++ case 4: ++ out += sprintf (out, "\\u%02x%02x%02x", ++ ((in[0] & 0x07) << 6) | ((in[1] & 0x3c) >> 2), ++ ((in[1] & 0x03) << 6) | ((in[2] & 0x3c) >> 2), ++ ((in[2] & 0x03) << 6) | ((in[3] & 0x3f))); ++ break; ++ default: ++ /* URG. */ ++ break; ++ } ++ ++ if (unicode_display == unicode_highlight && isatty (1)) ++ out += sprintf (out, "\033[0m"); /* Default colour. */ ++ break; ++ ++ default: ++ /* URG */ ++ break; ++ } ++ ++ * consumed = nchars; ++ return out - orig_out; ++ ++ invalid: ++ /* Not a valid UTF-8 sequence. */ ++ *out = *in; ++ * consumed = 1; ++ return 1; ++} ++ + /* Returns a version of IN with any control characters + replaced by escape sequences. Uses a static buffer +- if necessary. */ ++ if necessary. ++ ++ If unicode display is enabled, then also handles the ++ conversion of unicode characters. */ + + static const char * + sanitize_string (const char * in) +@@ -435,40 +563,50 @@ sanitize_string (const char * in) + of cases it will not be needed. */ + do + { +- char c = *in++; ++ unsigned char c = *in++; + + if (c == 0) + return original; + + if (ISCNTRL (c)) + break; ++ ++ if (unicode_display != unicode_default && c >= 0xc0) ++ break; + } + while (1); + + /* Copy the input, translating as needed. */ + in = original; +- if (buffer_len < (strlen (in) * 2)) ++ if (buffer_len < (strlen (in) * 9)) + { + free ((void *) buffer); +- buffer_len = strlen (in) * 2; ++ buffer_len = strlen (in) * 9; + buffer = xmalloc (buffer_len + 1); + } + + out = buffer; + do + { +- char c = *in++; ++ unsigned char c = *in++; + + if (c == 0) + break; + +- if (!ISCNTRL (c)) +- *out++ = c; +- else ++ if (ISCNTRL (c)) + { + *out++ = '^'; + *out++ = c + 0x40; + } ++ else if (unicode_display != unicode_default && c >= 0xc0) ++ { ++ unsigned int num_consumed; ++ ++ out += display_utf8 ((const unsigned char *)(in - 1), out, & num_consumed); ++ in += num_consumed - 1; ++ } ++ else ++ *out++ = c; + } + while (1); + +@@ -4481,6 +4619,24 @@ dump_symbols (bfd *abfd ATTRIBUTE_UNUSED + free (alloc); + } + } ++ else if (unicode_display != unicode_default ++ && name != NULL && *name != '\0') ++ { ++ const char * sanitized_name; ++ ++ /* If we want to sanitize the name, we do it here, and ++ temporarily clobber it while calling bfd_print_symbol. ++ FIXME: This is a gross hack. */ ++ sanitized_name = sanitize_string (name); ++ if (sanitized_name != name) ++ (*current)->name = sanitized_name; ++ else ++ sanitized_name = NULL; ++ bfd_print_symbol (cur_bfd, stdout, *current, ++ bfd_print_symbol_all); ++ if (sanitized_name != NULL) ++ (*current)->name = name; ++ } + else + bfd_print_symbol (cur_bfd, stdout, *current, + bfd_print_symbol_all); +@@ -5162,7 +5318,7 @@ main (int argc, char **argv) + set_default_bfd_target (); + + while ((c = getopt_long (argc, argv, +- "pP:ib:m:M:VvCdDlfFaHhrRtTxsSI:j:wE:zgeGW::", ++ "pP:ib:m:U:M:VvCdDlfFaHhrRtTxsSI:j:wE:zgeGW::", + long_options, (int *) 0)) + != EOF) + { +@@ -5441,6 +5597,23 @@ main (int argc, char **argv) + seenflag = TRUE; + break; + ++ case 'U': ++ if (streq (optarg, "default") || streq (optarg, "d")) ++ unicode_display = unicode_default; ++ else if (streq (optarg, "locale") || streq (optarg, "l")) ++ unicode_display = unicode_locale; ++ else if (streq (optarg, "escape") || streq (optarg, "e")) ++ unicode_display = unicode_escape; ++ else if (streq (optarg, "invalid") || streq (optarg, "i")) ++ unicode_display = unicode_invalid; ++ else if (streq (optarg, "hex") || streq (optarg, "x")) ++ unicode_display = unicode_hex; ++ else if (streq (optarg, "highlight") || streq (optarg, "h")) ++ unicode_display = unicode_highlight; ++ else ++ fatal (_("invalid argument to -U/--unicode: %s"), optarg); ++ break; ++ + case 'H': + usage (stdout, 0); + /* No need to set seenflag or to break - usage() does not return. */ +diff --git a/binutils/readelf.c b/binutils/readelf.c +--- a/binutils/readelf.c 2021-12-23 03:23:42.065783023 -0800 ++++ b/binutils/readelf.c 2021-12-23 03:30:36.582877519 -0800 +@@ -321,6 +321,19 @@ typedef enum print_mode + } + print_mode; + ++typedef enum unicode_display_type ++{ ++ unicode_default = 0, ++ unicode_locale, ++ unicode_escape, ++ unicode_hex, ++ unicode_highlight, ++ unicode_invalid ++} unicode_display_type; ++ ++static unicode_display_type unicode_display = unicode_default; ++ ++ + /* Versioned symbol info. */ + enum versioned_symbol_info + { +@@ -613,11 +626,18 @@ print_symbol (signed int width, const ch + if (c == 0) + break; + +- /* Do not print control characters directly as they can affect terminal +- settings. Such characters usually appear in the names generated +- by the assembler for local labels. */ +- if (ISCNTRL (c)) ++ if (ISPRINT (c)) ++ { ++ putchar (c); ++ width_remaining --; ++ num_printed ++; ++ } ++ else if (ISCNTRL (c)) + { ++ /* Do not print control characters directly as they can affect terminal ++ settings. Such characters usually appear in the names generated ++ by the assembler for local labels. */ ++ + if (width_remaining < 2) + break; + +@@ -625,11 +645,137 @@ print_symbol (signed int width, const ch + width_remaining -= 2; + num_printed += 2; + } +- else if (ISPRINT (c)) ++ else if (c == 0x7f) + { +- putchar (c); +- width_remaining --; +- num_printed ++; ++ if (width_remaining < 5) ++ break; ++ printf ("<DEL>"); ++ width_remaining -= 5; ++ num_printed += 5; ++ } ++ else if (unicode_display != unicode_locale ++ && unicode_display != unicode_default) ++ { ++ /* Display unicode characters as something else. */ ++ unsigned char bytes[4]; ++ bfd_boolean is_utf8; ++ unsigned int nbytes; ++ ++ bytes[0] = c; ++ ++ if (bytes[0] < 0xc0) ++ { ++ nbytes = 1; ++ is_utf8 = FALSE; ++ } ++ else ++ { ++ bytes[1] = *symbol++; ++ ++ if ((bytes[1] & 0xc0) != 0x80) ++ { ++ is_utf8 = FALSE; ++ /* Do not consume this character. It may only ++ be the first byte in the sequence that was ++ corrupt. */ ++ --symbol; ++ nbytes = 1; ++ } ++ else if ((bytes[0] & 0x20) == 0) ++ { ++ is_utf8 = TRUE; ++ nbytes = 2; ++ } ++ else ++ { ++ bytes[2] = *symbol++; ++ ++ if ((bytes[2] & 0xc0) != 0x80) ++ { ++ is_utf8 = FALSE; ++ symbol -= 2; ++ nbytes = 1; ++ } ++ else if ((bytes[0] & 0x10) == 0) ++ { ++ is_utf8 = TRUE; ++ nbytes = 3; ++ } ++ else ++ { ++ bytes[3] = *symbol++; ++ ++ nbytes = 4; ++ ++ if ((bytes[3] & 0xc0) != 0x80) ++ { ++ is_utf8 = FALSE; ++ symbol -= 3; ++ nbytes = 1; ++ } ++ else ++ is_utf8 = TRUE; ++ } ++ } ++ } ++ ++ if (unicode_display == unicode_invalid) ++ is_utf8 = FALSE; ++ ++ if (unicode_display == unicode_hex || ! is_utf8) ++ { ++ unsigned int i; ++ ++ if (width_remaining < (nbytes * 2) + 2) ++ break; ++ ++ putchar (is_utf8 ? '<' : '{'); ++ printf ("0x"); ++ for (i = 0; i < nbytes; i++) ++ printf ("%02x", bytes[i]); ++ putchar (is_utf8 ? '>' : '}'); ++ } ++ else ++ { ++ if (unicode_display == unicode_highlight && isatty (1)) ++ printf ("\x1B[31;47m"); /* Red. */ ++ ++ switch (nbytes) ++ { ++ case 2: ++ if (width_remaining < 6) ++ break; ++ printf ("\\u%02x%02x", ++ (bytes[0] & 0x1c) >> 2, ++ ((bytes[0] & 0x03) << 6) | (bytes[1] & 0x3f)); ++ break; ++ case 3: ++ if (width_remaining < 6) ++ break; ++ printf ("\\u%02x%02x", ++ ((bytes[0] & 0x0f) << 4) | ((bytes[1] & 0x3c) >> 2), ++ ((bytes[1] & 0x03) << 6) | (bytes[2] & 0x3f)); ++ break; ++ case 4: ++ if (width_remaining < 8) ++ break; ++ printf ("\\u%02x%02x%02x", ++ ((bytes[0] & 0x07) << 6) | ((bytes[1] & 0x3c) >> 2), ++ ((bytes[1] & 0x03) << 6) | ((bytes[2] & 0x3c) >> 2), ++ ((bytes[2] & 0x03) << 6) | (bytes[3] & 0x3f)); ++ ++ break; ++ default: ++ /* URG. */ ++ break; ++ } ++ ++ if (unicode_display == unicode_highlight && isatty (1)) ++ printf ("\033[0m"); /* Default colour. */ ++ } ++ ++ if (bytes[nbytes - 1] == 0) ++ break; + } + else + { +@@ -4556,6 +4702,7 @@ static struct option options[] = + {"syms", no_argument, 0, 's'}, + {"silent-truncation",no_argument, 0, 'T'}, + {"section-details", no_argument, 0, 't'}, ++ {"unicode", required_argument, NULL, 'U'}, + {"unwind", no_argument, 0, 'u'}, + {"version-info", no_argument, 0, 'V'}, + {"version", no_argument, 0, 'v'}, +@@ -4609,6 +4756,11 @@ usage (FILE * stream) + --no-demangle Do not demangle low-level symbol names. (This is the default)\n\ + --recurse-limit Enable a demangling recursion limit. (This is the default)\n\ + --no-recurse-limit Disable a demangling recursion limit\n\ ++ -U[dlexhi] --unicode=[default|locale|escape|hex|highlight|invalid]\n\ ++ Display unicode characters as determined by the current locale\n\ ++ (default), escape sequences, \"<hex sequences>\", highlighted\n\ ++ escape sequences, or treat them as invalid and display as\n\ ++ \"{hex sequences}\"\n\ + -n --notes Display the core notes (if present)\n\ + -r --relocs Display the relocations (if present)\n\ + -u --unwind Display the unwind info (if present)\n\ +@@ -4749,7 +4901,7 @@ parse_args (struct dump_data *dumpdata, + usage (stderr); + + while ((c = getopt_long +- (argc, argv, "ACDHILNR:STVWacdeghi:lnp:rstuvw::x:z", options, NULL)) != EOF) ++ (argc, argv, "ACDHILNR:STU:VWacdeghi:lnp:rstuvw::x:z", options, NULL)) != EOF) + { + switch (c) + { +@@ -4943,6 +5095,24 @@ parse_args (struct dump_data *dumpdata, + case OPTION_WITH_SYMBOL_VERSIONS: + /* Ignored for backward compatibility. */ + break; ++ case 'U': ++ if (optarg == NULL) ++ error (_("Missing arg to -U/--unicode")); /* Can this happen ? */ ++ else if (streq (optarg, "default") || streq (optarg, "d")) ++ unicode_display = unicode_default; ++ else if (streq (optarg, "locale") || streq (optarg, "l")) ++ unicode_display = unicode_locale; ++ else if (streq (optarg, "escape") || streq (optarg, "e")) ++ unicode_display = unicode_escape; ++ else if (streq (optarg, "invalid") || streq (optarg, "i")) ++ unicode_display = unicode_invalid; ++ else if (streq (optarg, "hex") || streq (optarg, "x")) ++ unicode_display = unicode_hex; ++ else if (streq (optarg, "highlight") || streq (optarg, "h")) ++ unicode_display = unicode_highlight; ++ else ++ error (_("invalid argument to -U/--unicode: %s"), optarg); ++ break; + + default: + /* xgettext:c-format */ +diff --git a/binutils/strings.c b/binutils/strings.c +--- a/binutils/strings.c 2021-12-23 03:23:38.485842662 -0800 ++++ b/binutils/strings.c 2021-12-23 03:30:36.586877452 -0800 +@@ -55,6 +55,19 @@ + -T {bfdname} + Specify a non-default object file format. + ++ --unicode={default|locale|invalid|hex|escape|highlight} ++ -u {d|l|i|x|e|h} ++ Determine how to handle UTF-8 unicode characters. The default ++ is no special treatment. All other versions of this option ++ only apply if the encoding is valid and enabling the option ++ implies --encoding=S. ++ The 'locale' option displays the characters according to the ++ current locale. The 'invalid' option treats them as ++ non-string characters. The 'hex' option displays them as hex ++ byte sequences. The 'escape' option displays them as escape ++ sequences and the 'highlight' option displays them as ++ coloured escape sequences. ++ + --output-separator=sep_string + -s sep_string String used to separate parsed strings in output. + Default is newline. +@@ -76,6 +89,22 @@ + #include "safe-ctype.h" + #include "bucomm.h" + ++#ifndef streq ++#define streq(a,b) (strcmp ((a),(b)) == 0) ++#endif ++ ++typedef enum unicode_display_type ++{ ++ unicode_default = 0, ++ unicode_locale, ++ unicode_escape, ++ unicode_hex, ++ unicode_highlight, ++ unicode_invalid ++} unicode_display_type; ++ ++static unicode_display_type unicode_display = unicode_default; ++ + #define STRING_ISGRAPHIC(c) \ + ( (c) >= 0 \ + && (c) <= 255 \ +@@ -94,7 +123,7 @@ extern int errno; + static int address_radix; + + /* Minimum length of sequence of graphic chars to trigger output. */ +-static int string_min; ++static unsigned int string_min; + + /* Whether or not we include all whitespace as a graphic char. */ + static bfd_boolean include_all_whitespace; +@@ -121,21 +150,22 @@ static char *output_separator; + static struct option long_options[] = + { + {"all", no_argument, NULL, 'a'}, ++ {"bytes", required_argument, NULL, 'n'}, + {"data", no_argument, NULL, 'd'}, ++ {"encoding", required_argument, NULL, 'e'}, ++ {"help", no_argument, NULL, 'h'}, ++ {"include-all-whitespace", no_argument, NULL, 'w'}, ++ {"output-separator", required_argument, NULL, 's'}, + {"print-file-name", no_argument, NULL, 'f'}, +- {"bytes", required_argument, NULL, 'n'}, + {"radix", required_argument, NULL, 't'}, +- {"include-all-whitespace", no_argument, NULL, 'w'}, +- {"encoding", required_argument, NULL, 'e'}, + {"target", required_argument, NULL, 'T'}, +- {"output-separator", required_argument, NULL, 's'}, +- {"help", no_argument, NULL, 'h'}, ++ {"unicode", required_argument, NULL, 'U'}, + {"version", no_argument, NULL, 'v'}, + {NULL, 0, NULL, 0} + }; + + static bfd_boolean strings_file (char *); +-static void print_strings (const char *, FILE *, file_ptr, int, int, char *); ++static void print_strings (const char *, FILE *, file_ptr, int, char *); + static void usage (FILE *, int) ATTRIBUTE_NORETURN; + + int main (int, char **); +@@ -173,7 +203,7 @@ main (int argc, char **argv) + encoding = 's'; + output_separator = NULL; + +- while ((optc = getopt_long (argc, argv, "adfhHn:wot:e:T:s:Vv0123456789", ++ while ((optc = getopt_long (argc, argv, "adfhHn:wot:e:T:s:U:Vv0123456789", + long_options, (int *) 0)) != EOF) + { + switch (optc) +@@ -244,7 +274,24 @@ main (int argc, char **argv) + + case 's': + output_separator = optarg; +- break; ++ break; ++ ++ case 'U': ++ if (streq (optarg, "default") || streq (optarg, "d")) ++ unicode_display = unicode_default; ++ else if (streq (optarg, "locale") || streq (optarg, "l")) ++ unicode_display = unicode_locale; ++ else if (streq (optarg, "escape") || streq (optarg, "e")) ++ unicode_display = unicode_escape; ++ else if (streq (optarg, "invalid") || streq (optarg, "i")) ++ unicode_display = unicode_invalid; ++ else if (streq (optarg, "hex") || streq (optarg, "x")) ++ unicode_display = unicode_hex; ++ else if (streq (optarg, "highlight") || streq (optarg, "h")) ++ unicode_display = unicode_highlight; ++ else ++ fatal (_("invalid argument to -U/--unicode: %s"), optarg); ++ break; + + case 'V': + case 'v': +@@ -260,6 +307,9 @@ main (int argc, char **argv) + } + } + ++ if (unicode_display != unicode_default) ++ encoding = 'S'; ++ + if (numeric_opt != 0) + { + string_min = (int) strtoul (argv[numeric_opt - 1] + 1, &s, 0); +@@ -295,14 +345,14 @@ main (int argc, char **argv) + { + datasection_only = FALSE; + SET_BINARY (fileno (stdin)); +- print_strings ("{standard input}", stdin, 0, 0, 0, (char *) NULL); ++ print_strings ("{standard input}", stdin, 0, 0, (char *) NULL); + files_given = TRUE; + } + else + { + for (; optind < argc; ++optind) + { +- if (strcmp (argv[optind], "-") == 0) ++ if (streq (argv[optind], "-")) + datasection_only = FALSE; + else + { +@@ -344,7 +394,7 @@ strings_a_section (bfd *abfd, asection * + } + + *got_a_section = TRUE; +- print_strings (filename, NULL, sect->filepos, 0, sectsize, (char *) mem); ++ print_strings (filename, NULL, sect->filepos, sectsize, (char *) mem); + free (mem); + } + +@@ -429,7 +479,7 @@ strings_file (char *file) + return FALSE; + } + +- print_strings (file, stream, (file_ptr) 0, 0, 0, (char *) 0); ++ print_strings (file, stream, (file_ptr) 0, 0, (char *) NULL); + + if (fclose (stream) == EOF) + { +@@ -553,11 +603,632 @@ unget_part_char (long c, file_ptr *addre + } + } + } ++ ++ ++static void ++print_filename_and_address (const char * filename, file_ptr address) ++{ ++ if (print_filenames) ++ printf ("%s: ", filename); ++ ++ if (! print_addresses) ++ return; ++ ++ switch (address_radix) ++ { ++ case 8: ++ if (sizeof (address) > sizeof (long)) ++ { ++#ifndef __MSVCRT__ ++ printf ("%7llo ", (unsigned long long) address); ++#else ++ printf ("%7I64o ", (unsigned long long) address); ++#endif ++ } ++ else ++ printf ("%7lo ", (unsigned long) address); ++ break; ++ ++ case 10: ++ if (sizeof (address) > sizeof (long)) ++ { ++#ifndef __MSVCRT__ ++ printf ("%7llu ", (unsigned long long) address); ++#else ++ printf ("%7I64d ", (unsigned long long) address); ++#endif ++ } ++ else ++ printf ("%7ld ", (long) address); ++ break; ++ ++ case 16: ++ if (sizeof (address) > sizeof (long)) ++ { ++#ifndef __MSVCRT__ ++ printf ("%7llx ", (unsigned long long) address); ++#else ++ printf ("%7I64x ", (unsigned long long) address); ++#endif ++ } ++ else ++ printf ("%7lx ", (unsigned long) address); ++ break; ++ } ++} ++ ++/* Return non-zero if the bytes starting at BUFFER form a valid UTF-8 encoding. ++ If the encoding is valid then returns the number of bytes it uses. */ ++ ++static unsigned int ++is_valid_utf8 (const unsigned char * buffer, unsigned long buflen) ++{ ++ if (buffer[0] < 0xc0) ++ return 0; ++ ++ if (buflen < 2) ++ return 0; ++ ++ if ((buffer[1] & 0xc0) != 0x80) ++ return 0; ++ ++ if ((buffer[0] & 0x20) == 0) ++ return 2; ++ ++ if (buflen < 3) ++ return 0; ++ ++ if ((buffer[2] & 0xc0) != 0x80) ++ return 0; ++ ++ if ((buffer[0] & 0x10) == 0) ++ return 3; ++ ++ if (buflen < 4) ++ return 0; ++ ++ if ((buffer[3] & 0xc0) != 0x80) ++ return 0; ++ ++ return 4; ++} ++ ++/* Display a UTF-8 encoded character in BUFFER according to the setting ++ of unicode_display. The character is known to be valid. ++ Returns the number of bytes consumed. */ ++ ++static unsigned int ++display_utf8_char (const unsigned char * buffer) ++{ ++ unsigned int j; ++ unsigned int utf8_len; ++ ++ switch (buffer[0] & 0x30) ++ { ++ case 0x00: ++ case 0x10: ++ utf8_len = 2; ++ break; ++ case 0x20: ++ utf8_len = 3; ++ break; ++ default: ++ utf8_len = 4; ++ } ++ ++ switch (unicode_display) ++ { ++ default: ++ fprintf (stderr, "ICE: unexpected unicode display type\n"); ++ break; ++ ++ case unicode_escape: ++ case unicode_highlight: ++ if (unicode_display == unicode_highlight && isatty (1)) ++ printf ("\x1B[31;47m"); /* Red. */ ++ ++ switch (utf8_len) ++ { ++ case 2: ++ printf ("\\u%02x%02x", ++ ((buffer[0] & 0x1c) >> 2), ++ ((buffer[0] & 0x03) << 6) | (buffer[1] & 0x3f)); ++ break; ++ ++ case 3: ++ printf ("\\u%02x%02x", ++ ((buffer[0] & 0x0f) << 4) | ((buffer[1] & 0x3c) >> 2), ++ ((buffer[1] & 0x03) << 6) | ((buffer[2] & 0x3f))); ++ break; ++ ++ case 4: ++ printf ("\\u%02x%02x%02x", ++ ((buffer[0] & 0x07) << 6) | ((buffer[1] & 0x3c) >> 2), ++ ((buffer[1] & 0x03) << 6) | ((buffer[2] & 0x3c) >> 2), ++ ((buffer[2] & 0x03) << 6) | ((buffer[3] & 0x3f))); ++ break; ++ default: ++ /* URG. */ ++ break; ++ } ++ ++ if (unicode_display == unicode_highlight && isatty (1)) ++ printf ("\033[0m"); /* Default colour. */ ++ break; ++ ++ case unicode_hex: ++ putchar ('<'); ++ printf ("0x"); ++ for (j = 0; j < utf8_len; j++) ++ printf ("%02x", buffer [j]); ++ putchar ('>'); ++ break; ++ ++ case unicode_locale: ++ printf ("%.1s", buffer); ++ break; ++ } ++ ++ return utf8_len; ++} ++ ++/* Display strings in BUFFER. Treat any UTF-8 encoded characters encountered ++ according to the setting of the unicode_display variable. The buffer ++ contains BUFLEN bytes. ++ ++ Display the characters as if they started at ADDRESS and are contained in ++ FILENAME. */ ++ ++static void ++print_unicode_buffer (const char * filename, ++ file_ptr address, ++ const unsigned char * buffer, ++ unsigned long buflen) ++{ ++ /* Paranoia checks... */ ++ if (filename == NULL ++ || buffer == NULL ++ || unicode_display == unicode_default ++ || encoding != 'S' ++ || encoding_bytes != 1) ++ { ++ fprintf (stderr, "ICE: bad arguments to print_unicode_buffer\n"); ++ return; ++ } ++ ++ if (buflen == 0) ++ return; ++ ++ /* We must only display strings that are at least string_min *characters* ++ long. So we scan the buffer in two stages. First we locate the start ++ of a potential string. Then we walk along it until we have found ++ string_min characters. Then we go back to the start point and start ++ displaying characters according to the unicode_display setting. */ ++ ++ unsigned long start_point = 0; ++ unsigned long i = 0; ++ unsigned int char_len = 1; ++ unsigned int num_found = 0; ++ ++ for (i = 0; i < buflen; i += char_len) ++ { ++ int c = buffer[i]; ++ ++ char_len = 1; ++ ++ /* Find the first potential character of a string. */ ++ if (! STRING_ISGRAPHIC (c)) ++ { ++ num_found = 0; ++ continue; ++ } ++ ++ if (c > 126) ++ { ++ if (c < 0xc0) ++ { ++ num_found = 0; ++ continue; ++ } ++ ++ if ((char_len = is_valid_utf8 (buffer + i, buflen - i)) == 0) ++ { ++ char_len = 1; ++ num_found = 0; ++ continue; ++ } ++ ++ if (unicode_display == unicode_invalid) ++ { ++ /* We have found a valid UTF-8 character, but we treat it as non-graphic. */ ++ num_found = 0; ++ continue; ++ } ++ } ++ ++ if (num_found == 0) ++ /* We have found a potential starting point for a string. */ ++ start_point = i; ++ ++ ++ num_found; ++ ++ if (num_found >= string_min) ++ break; ++ } ++ ++ if (num_found < string_min) ++ return; ++ ++ print_filename_and_address (filename, address + start_point); ++ ++ /* We have found string_min characters. Display them and any ++ more that follow. */ ++ for (i = start_point; i < buflen; i += char_len) ++ { ++ int c = buffer[i]; ++ ++ char_len = 1; ++ ++ if (! STRING_ISGRAPHIC (c)) ++ break; ++ else if (c < 127) ++ putchar (c); ++ else if (! is_valid_utf8 (buffer + i, buflen - i)) ++ break; ++ else if (unicode_display == unicode_invalid) ++ break; ++ else ++ char_len = display_utf8_char (buffer + i); ++ } ++ ++ if (output_separator) ++ fputs (output_separator, stdout); ++ else ++ putchar ('\n'); ++ ++ /* FIXME: Using tail recursion here is lazy programming... */ ++ print_unicode_buffer (filename, address + i, buffer + i, buflen - i); ++} ++ ++static int ++get_unicode_byte (FILE * stream, ++ unsigned char * putback, ++ unsigned int * num_putback, ++ unsigned int * num_read) ++{ ++ if (* num_putback > 0) ++ { ++ * num_putback = * num_putback - 1; ++ return putback [* num_putback]; ++ } ++ ++ * num_read = * num_read + 1; ++ ++#if defined(HAVE_GETC_UNLOCKED) && HAVE_DECL_GETC_UNLOCKED ++ return getc_unlocked (stream); ++#else ++ return getc (stream); ++#endif ++} ++ ++/* Helper function for print_unicode_stream. */ ++ ++static void ++print_unicode_stream_body (const char * filename, ++ file_ptr address, ++ FILE * stream, ++ unsigned char * putback_buf, ++ unsigned int num_putback, ++ unsigned char * print_buf) ++{ ++ /* It would be nice if we could just read the stream into a buffer ++ and then process if with print_unicode_buffer. But the input ++ might be huge or it might time-locked (eg stdin). So instead ++ we go one byte at a time... */ ++ ++ file_ptr start_point = 0; ++ unsigned int num_read = 0; ++ unsigned int num_chars = 0; ++ unsigned int num_print = 0; ++ int c; ++ ++ /* Find a series of string_min characters. Put them into print_buf. */ ++ do ++ { ++ if (num_chars >= string_min) ++ break; ++ ++ c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); ++ if (c == EOF) ++ break; ++ ++ if (! STRING_ISGRAPHIC (c)) ++ { ++ num_chars = num_print = 0; ++ continue; ++ } ++ ++ if (num_chars == 0) ++ start_point = num_read - 1; ++ ++ if (c < 127) ++ { ++ print_buf[num_print] = c; ++ num_chars ++; ++ num_print ++; ++ continue; ++ } ++ ++ if (c < 0xc0) ++ { ++ num_chars = num_print = 0; ++ continue; ++ } ++ ++ /* We *might* have a UTF-8 sequence. Time to start peeking. */ ++ char utf8[4]; ++ ++ utf8[0] = c; ++ c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); ++ if (c == EOF) ++ break; ++ utf8[1] = c; ++ ++ if ((utf8[1] & 0xc0) != 0x80) ++ { ++ /* Invalid UTF-8. */ ++ putback_buf[num_putback++] = utf8[1]; ++ num_chars = num_print = 0; ++ continue; ++ } ++ else if ((utf8[0] & 0x20) == 0) ++ { ++ /* A valid 2-byte UTF-8 encoding. */ ++ if (unicode_display == unicode_invalid) ++ { ++ putback_buf[num_putback++] = utf8[1]; ++ num_chars = num_print = 0; ++ } ++ else ++ { ++ print_buf[num_print ++] = utf8[0]; ++ print_buf[num_print ++] = utf8[1]; ++ num_chars ++; ++ } ++ continue; ++ } ++ ++ c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); ++ if (c == EOF) ++ break; ++ utf8[2] = c; ++ ++ if ((utf8[2] & 0xc0) != 0x80) ++ { ++ /* Invalid UTF-8. */ ++ putback_buf[num_putback++] = utf8[2]; ++ putback_buf[num_putback++] = utf8[1]; ++ num_chars = num_print = 0; ++ continue; ++ } ++ else if ((utf8[0] & 0x10) == 0) ++ { ++ /* A valid 3-byte UTF-8 encoding. */ ++ if (unicode_display == unicode_invalid) ++ { ++ putback_buf[num_putback++] = utf8[2]; ++ putback_buf[num_putback++] = utf8[1]; ++ num_chars = num_print = 0; ++ } ++ else ++ { ++ print_buf[num_print ++] = utf8[0]; ++ print_buf[num_print ++] = utf8[1]; ++ print_buf[num_print ++] = utf8[2]; ++ num_chars ++; ++ } ++ continue; ++ } ++ ++ c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); ++ if (c == EOF) ++ break; ++ utf8[3] = c; ++ ++ if ((utf8[3] & 0xc0) != 0x80) ++ { ++ /* Invalid UTF-8. */ ++ putback_buf[num_putback++] = utf8[3]; ++ putback_buf[num_putback++] = utf8[2]; ++ putback_buf[num_putback++] = utf8[1]; ++ num_chars = num_print = 0; ++ } ++ /* We have a valid 4-byte UTF-8 encoding. */ ++ else if (unicode_display == unicode_invalid) ++ { ++ putback_buf[num_putback++] = utf8[3]; ++ putback_buf[num_putback++] = utf8[1]; ++ putback_buf[num_putback++] = utf8[2]; ++ num_chars = num_print = 0; ++ } ++ else ++ { ++ print_buf[num_print ++] = utf8[0]; ++ print_buf[num_print ++] = utf8[1]; ++ print_buf[num_print ++] = utf8[2]; ++ print_buf[num_print ++] = utf8[3]; ++ num_chars ++; ++ } ++ } ++ while (1); ++ ++ if (num_chars >= string_min) ++ { ++ /* We know that we have string_min valid characters in print_buf, ++ and there may be more to come in the stream. Start displaying ++ them. */ ++ ++ print_filename_and_address (filename, address + start_point); ++ ++ unsigned int i; ++ for (i = 0; i < num_print;) ++ { ++ if (print_buf[i] < 127) ++ putchar (print_buf[i++]); ++ else ++ i += display_utf8_char (print_buf + i); ++ } ++ ++ /* OK so now we have to start read unchecked bytes. */ ++ ++ /* Find a series of string_min characters. Put them into print_buf. */ ++ do ++ { ++ c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); ++ if (c == EOF) ++ break; ++ ++ if (! STRING_ISGRAPHIC (c)) ++ break; ++ ++ if (c < 127) ++ { ++ putchar (c); ++ continue; ++ } ++ ++ if (c < 0xc0) ++ break; ++ ++ /* We *might* have a UTF-8 sequence. Time to start peeking. */ ++ unsigned char utf8[4]; ++ ++ utf8[0] = c; ++ c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); ++ if (c == EOF) ++ break; ++ utf8[1] = c; ++ ++ if ((utf8[1] & 0xc0) != 0x80) ++ { ++ /* Invalid UTF-8. */ ++ putback_buf[num_putback++] = utf8[1]; ++ break; ++ } ++ else if ((utf8[0] & 0x20) == 0) ++ { ++ /* Valid 2-byte UTF-8. */ ++ if (unicode_display == unicode_invalid) ++ { ++ putback_buf[num_putback++] = utf8[1]; ++ break; ++ } ++ else ++ { ++ (void) display_utf8_char (utf8); ++ continue; ++ } ++ } ++ ++ c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); ++ if (c == EOF) ++ break; ++ utf8[2] = c; ++ ++ if ((utf8[2] & 0xc0) != 0x80) ++ { ++ /* Invalid UTF-8. */ ++ putback_buf[num_putback++] = utf8[2]; ++ putback_buf[num_putback++] = utf8[1]; ++ break; ++ } ++ else if ((utf8[0] & 0x10) == 0) ++ { ++ /* Valid 3-byte UTF-8. */ ++ if (unicode_display == unicode_invalid) ++ { ++ putback_buf[num_putback++] = utf8[2]; ++ putback_buf[num_putback++] = utf8[1]; ++ break; ++ } ++ else ++ { ++ (void) display_utf8_char (utf8); ++ continue; ++ } ++ } ++ ++ c = get_unicode_byte (stream, putback_buf, & num_putback, & num_read); ++ if (c == EOF) ++ break; ++ utf8[3] = c; ++ ++ if ((utf8[3] & 0xc0) != 0x80) ++ { ++ /* Invalid UTF-8. */ ++ putback_buf[num_putback++] = utf8[3]; ++ putback_buf[num_putback++] = utf8[2]; ++ putback_buf[num_putback++] = utf8[1]; ++ break; ++ } ++ else if (unicode_display == unicode_invalid) ++ { ++ putback_buf[num_putback++] = utf8[3]; ++ putback_buf[num_putback++] = utf8[2]; ++ putback_buf[num_putback++] = utf8[1]; ++ break; ++ } ++ else ++ /* A valid 4-byte UTF-8 encoding. */ ++ (void) display_utf8_char (utf8); ++ } ++ while (1); ++ ++ if (output_separator) ++ fputs (output_separator, stdout); ++ else ++ putchar ('\n'); ++ } ++ ++ if (c != EOF) ++ /* FIXME: Using tail recursion here is lazy, but it works. */ ++ print_unicode_stream_body (filename, address + num_read, stream, putback_buf, num_putback, print_buf); ++} ++ ++/* Display strings read in from STREAM. Treat any UTF-8 encoded characters ++ encountered according to the setting of the unicode_display variable. ++ The stream is positioned at ADDRESS and is attached to FILENAME. */ ++ ++static void ++print_unicode_stream (const char * filename, ++ file_ptr address, ++ FILE * stream) ++{ ++ /* Paranoia checks... */ ++ if (filename == NULL ++ || stream == NULL ++ || unicode_display == unicode_default ++ || encoding != 'S' ++ || encoding_bytes != 1) ++ { ++ fprintf (stderr, "ICE: bad arguments to print_unicode_stream\n"); ++ return; ++ } ++ ++ /* Allocate space for string_min 4-byte utf-8 characters. */ ++ unsigned char * print_buf = xmalloc ((4 * string_min) + 1); ++ /* We should never have to put back more than 4 bytes. */ ++ unsigned char putback_buf[5]; ++ unsigned int num_putback = 0; ++ ++ print_unicode_stream_body (filename, address, stream, putback_buf, num_putback, print_buf); ++ free (print_buf); ++} ++ + + /* Find the strings in file FILENAME, read from STREAM. + Assume that STREAM is positioned so that the next byte read + is at address ADDRESS in the file. +- Stop reading at address STOP_POINT in the file, if nonzero. + + If STREAM is NULL, do not read from it. + The caller can supply a buffer of characters +@@ -568,20 +1239,29 @@ unget_part_char (long c, file_ptr *addre + + static void + print_strings (const char *filename, FILE *stream, file_ptr address, +- int stop_point, int magiccount, char *magic) ++ int magiccount, char *magic) + { ++ if (unicode_display != unicode_default) ++ { ++ if (magic != NULL) ++ print_unicode_buffer (filename, address, ++ (const unsigned char *) magic, magiccount); ++ ++ if (stream != NULL) ++ print_unicode_stream (filename, address, stream); ++ return; ++ } ++ + char *buf = (char *) xmalloc (sizeof (char) * (string_min + 1)); + + while (1) + { + file_ptr start; +- int i; ++ unsigned int i; + long c; + + /* See if the next `string_min' chars are all graphic chars. */ + tryline: +- if (stop_point && address >= stop_point) +- break; + start = address; + for (i = 0; i < string_min; i++) + { +@@ -604,69 +1284,7 @@ print_strings (const char *filename, FIL + /* We found a run of `string_min' graphic characters. Print up + to the next non-graphic character. */ + +- if (print_filenames) +- printf ("%s: ", filename); +- if (print_addresses) +- switch (address_radix) +- { +- case 8: +-#ifdef HAVE_LONG_LONG +- if (sizeof (start) > sizeof (long)) +- { +-# ifndef __MSVCRT__ +- printf ("%7llo ", (unsigned long long) start); +-# else +- printf ("%7I64o ", (unsigned long long) start); +-# endif +- } +- else +-#elif !BFD_HOST_64BIT_LONG +- if (start != (unsigned long) start) +- printf ("++%7lo ", (unsigned long) start); +- else +-#endif +- printf ("%7lo ", (unsigned long) start); +- break; +- +- case 10: +-#ifdef HAVE_LONG_LONG +- if (sizeof (start) > sizeof (long)) +- { +-# ifndef __MSVCRT__ +- printf ("%7llu ", (unsigned long long) start); +-# else +- printf ("%7I64d ", (unsigned long long) start); +-# endif +- } +- else +-#elif !BFD_HOST_64BIT_LONG +- if (start != (unsigned long) start) +- printf ("++%7lu ", (unsigned long) start); +- else +-#endif +- printf ("%7ld ", (long) start); +- break; +- +- case 16: +-#ifdef HAVE_LONG_LONG +- if (sizeof (start) > sizeof (long)) +- { +-# ifndef __MSVCRT__ +- printf ("%7llx ", (unsigned long long) start); +-# else +- printf ("%7I64x ", (unsigned long long) start); +-# endif +- } +- else +-#elif !BFD_HOST_64BIT_LONG +- if (start != (unsigned long) start) +- printf ("%lx%8.8lx ", (unsigned long) (start >> 32), +- (unsigned long) (start & 0xffffffff)); +- else +-#endif +- printf ("%7lx ", (unsigned long) start); +- break; +- } ++ print_filename_and_address (filename, start); + + buf[i] = '\0'; + fputs (buf, stdout); +@@ -718,6 +1336,8 @@ usage (FILE *stream, int status) + -T --target=<BFDNAME> Specify the binary file format\n\ + -e --encoding={s,S,b,l,B,L} Select character size and endianness:\n\ + s = 7-bit, S = 8-bit, {b,l} = 16-bit, {B,L} = 32-bit\n\ ++ --unicode={default|show|invalid|hex|escape|highlight}\n\ ++ -u {d|s|i|x|e|h} Specify how to treat UTF-8 encoded unicode characters\n\ + -s --output-separator=<string> String used to separate strings in output.\n\ + @<file> Read options from <file>\n\ + -h --help Display this information\n\ diff --git a/meta/recipes-devtools/binutils/binutils/0001-CVE-2021-45078.patch b/meta/recipes-devtools/binutils/binutils/0001-CVE-2021-45078.patch new file mode 100644 index 0000000000..f118e2599b --- /dev/null +++ b/meta/recipes-devtools/binutils/binutils/0001-CVE-2021-45078.patch @@ -0,0 +1,255 @@ +From 161e87d12167b1e36193385485c1f6ce92f74f02 Mon Sep 17 00:00:00 2001 +From: Alan Modra <amodra@gmail.com> +Date: Wed, 15 Dec 2021 11:48:42 +1030 +Subject: [PATCH] PR28694, Out-of-bounds write in stab_xcoff_builtin_type + + PR 28694 + * stabs.c (stab_xcoff_builtin_type): Make typenum unsigned. + Negate typenum earlier, simplifying bounds checking. Correct + off-by-one indexing. Adjust switch cases. + + +CVE: CVE-2021-45078 +Upstream-Status: Backport [https://sourceware.org/git/?p=binutils-gdb.git;a=patch;h=161e87d12167b1e36193385485c1f6ce92f74f02] + +Signed-off-by: Sundeep KOKKONDA <sundeep.kokkonda@gmail.com> +--- + binutils/stabs.c | 87 ++++++++++++++++++++++++------------------------ + 1 file changed, 43 insertions(+), 44 deletions(-) + + +diff --git a/binutils/stabs.c b/binutils/stabs.c +index 274bfb0e7fa..83ee3ea5fa4 100644 +--- a/binutils/stabs.c ++++ b/binutils/stabs.c +@@ -202,7 +202,7 @@ static debug_type stab_find_type (void *, struct stab_handle *, const int *); + static bfd_boolean stab_record_type + (void *, struct stab_handle *, const int *, debug_type); + static debug_type stab_xcoff_builtin_type +- (void *, struct stab_handle *, int); ++ (void *, struct stab_handle *, unsigned int); + static debug_type stab_find_tagged_type + (void *, struct stab_handle *, const char *, int, enum debug_type_kind); + static debug_type *stab_demangle_argtypes +@@ -3496,166 +3496,167 @@ stab_record_type (void *dhandle ATTRIBUTE_UNUSED, struct stab_handle *info, + + static debug_type + stab_xcoff_builtin_type (void *dhandle, struct stab_handle *info, +- int typenum) ++ unsigned int typenum) + { + debug_type rettype; + const char *name; + +- if (typenum >= 0 || typenum < -XCOFF_TYPE_COUNT) ++ typenum = -typenum - 1; ++ if (typenum >= XCOFF_TYPE_COUNT) + { +- fprintf (stderr, _("Unrecognized XCOFF type %d\n"), typenum); ++ fprintf (stderr, _("Unrecognized XCOFF type %d\n"), -typenum - 1); + return DEBUG_TYPE_NULL; + } +- if (info->xcoff_types[-typenum] != NULL) +- return info->xcoff_types[-typenum]; ++ if (info->xcoff_types[typenum] != NULL) ++ return info->xcoff_types[typenum]; + +- switch (-typenum) ++ switch (typenum) + { +- case 1: ++ case 0: + /* The size of this and all the other types are fixed, defined + by the debugging format. */ + name = "int"; + rettype = debug_make_int_type (dhandle, 4, FALSE); + break; +- case 2: ++ case 1: + name = "char"; + rettype = debug_make_int_type (dhandle, 1, FALSE); + break; +- case 3: ++ case 2: + name = "short"; + rettype = debug_make_int_type (dhandle, 2, FALSE); + break; +- case 4: ++ case 3: + name = "long"; + rettype = debug_make_int_type (dhandle, 4, FALSE); + break; +- case 5: ++ case 4: + name = "unsigned char"; + rettype = debug_make_int_type (dhandle, 1, TRUE); + break; +- case 6: ++ case 5: + name = "signed char"; + rettype = debug_make_int_type (dhandle, 1, FALSE); + break; +- case 7: ++ case 6: + name = "unsigned short"; + rettype = debug_make_int_type (dhandle, 2, TRUE); + break; +- case 8: ++ case 7: + name = "unsigned int"; + rettype = debug_make_int_type (dhandle, 4, TRUE); + break; +- case 9: ++ case 8: + name = "unsigned"; + rettype = debug_make_int_type (dhandle, 4, TRUE); + break; +- case 10: ++ case 9: + name = "unsigned long"; + rettype = debug_make_int_type (dhandle, 4, TRUE); + break; +- case 11: ++ case 10: + name = "void"; + rettype = debug_make_void_type (dhandle); + break; +- case 12: ++ case 11: + /* IEEE single precision (32 bit). */ + name = "float"; + rettype = debug_make_float_type (dhandle, 4); + break; +- case 13: ++ case 12: + /* IEEE double precision (64 bit). */ + name = "double"; + rettype = debug_make_float_type (dhandle, 8); + break; +- case 14: ++ case 13: + /* This is an IEEE double on the RS/6000, and different machines + with different sizes for "long double" should use different + negative type numbers. See stabs.texinfo. */ + name = "long double"; + rettype = debug_make_float_type (dhandle, 8); + break; +- case 15: ++ case 14: + name = "integer"; + rettype = debug_make_int_type (dhandle, 4, FALSE); + break; +- case 16: ++ case 15: + name = "boolean"; + rettype = debug_make_bool_type (dhandle, 4); + break; +- case 17: ++ case 16: + name = "short real"; + rettype = debug_make_float_type (dhandle, 4); + break; +- case 18: ++ case 17: + name = "real"; + rettype = debug_make_float_type (dhandle, 8); + break; +- case 19: ++ case 18: + /* FIXME */ + name = "stringptr"; + rettype = NULL; + break; +- case 20: ++ case 19: + /* FIXME */ + name = "character"; + rettype = debug_make_int_type (dhandle, 1, TRUE); + break; +- case 21: ++ case 20: + name = "logical*1"; + rettype = debug_make_bool_type (dhandle, 1); + break; +- case 22: ++ case 21: + name = "logical*2"; + rettype = debug_make_bool_type (dhandle, 2); + break; +- case 23: ++ case 22: + name = "logical*4"; + rettype = debug_make_bool_type (dhandle, 4); + break; +- case 24: ++ case 23: + name = "logical"; + rettype = debug_make_bool_type (dhandle, 4); + break; +- case 25: ++ case 24: + /* Complex type consisting of two IEEE single precision values. */ + name = "complex"; + rettype = debug_make_complex_type (dhandle, 8); + break; +- case 26: ++ case 25: + /* Complex type consisting of two IEEE double precision values. */ + name = "double complex"; + rettype = debug_make_complex_type (dhandle, 16); + break; +- case 27: ++ case 26: + name = "integer*1"; + rettype = debug_make_int_type (dhandle, 1, FALSE); + break; +- case 28: ++ case 27: + name = "integer*2"; + rettype = debug_make_int_type (dhandle, 2, FALSE); + break; +- case 29: ++ case 28: + name = "integer*4"; + rettype = debug_make_int_type (dhandle, 4, FALSE); + break; +- case 30: ++ case 29: + /* FIXME */ + name = "wchar"; + rettype = debug_make_int_type (dhandle, 2, FALSE); + break; +- case 31: ++ case 30: + name = "long long"; + rettype = debug_make_int_type (dhandle, 8, FALSE); + break; +- case 32: ++ case 31: + name = "unsigned long long"; + rettype = debug_make_int_type (dhandle, 8, TRUE); + break; +- case 33: ++ case 32: + name = "logical*8"; + rettype = debug_make_bool_type (dhandle, 8); + break; +- case 34: ++ case 33: + name = "integer*8"; + rettype = debug_make_int_type (dhandle, 8, FALSE); + break; +@@ -3664,9 +3665,7 @@ stab_xcoff_builtin_type (void *dhandle, struct stab_handle *info, + } + + rettype = debug_name_type (dhandle, name, rettype); +- +- info->xcoff_types[-typenum] = rettype; +- ++ info->xcoff_types[typenum] = rettype; + return rettype; + } + +-- +2.27.0 + diff --git a/meta/recipes-devtools/binutils/binutils/0002-CVE-2021-20197.patch b/meta/recipes-devtools/binutils/binutils/0002-CVE-2021-20197.patch deleted file mode 100644 index 3771f571eb..0000000000 --- a/meta/recipes-devtools/binutils/binutils/0002-CVE-2021-20197.patch +++ /dev/null @@ -1,170 +0,0 @@ -From d3edaa91d4cf7202ec14342410194841e2f67f12 Mon Sep 17 00:00:00 2001 -From: Alan Modra <amodra@gmail.com> -Date: Fri, 26 Feb 2021 11:30:32 +1030 -Subject: [PATCH] Reinstate various pieces backed out from smart_rename changes - -In the interests of a stable release various last minute smart_rename -patches were backed out of the 2.36 branch. The main reason to -reinstate some of those backed out changes here is to make necessary -followup fixes to commit 8e03235147a9 simple cherry-picks from -mainline. A secondary reason is that ar -M support isn't fixed for -pr26945 without this patch. - - PR 26945 - * ar.c: Don't include libbfd.h. - (write_archive): Replace xmalloc+strcpy with xstrdup. - * arsup.c (temp_name, real_ofd): New static variables. - (ar_open): Use make_tempname and bfd_fdopenw. - (ar_save): Adjust to suit ar_open changes. - * objcopy.c: Don't include libbfd.h. - * rename.c: Rename and reorder variables. - -(cherry picked from commit 95b91a043aeaeb546d2fea556d84a2de1e917770) - -Upstream-Status: Backport [https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;h=d3edaa91d4cf7202ec14342410194841e2f67f12] -CVE: CVE-2021-20197 -Signed-off-by: Vinay Kumar <vinay.m.engg@gmail.com> ---- - binutils/ar.c | 4 +--- - binutils/arsup.c | 37 +++++++++++++++++++++++++------------ - binutils/objcopy.c | 1 - - binutils/rename.c | 6 +++--- - 5 files changed, 42 insertions(+), 19 deletions(-) - -diff --git a/binutils/ar.c b/binutils/ar.c -index 3a91708b51c..44df48c5c67 100644 ---- a/binutils/ar.c -+++ b/binutils/ar.c -@@ -25,7 +25,6 @@ - - #include "sysdep.h" - #include "bfd.h" --#include "libbfd.h" - #include "libiberty.h" - #include "progress.h" - #include "getopt.h" -@@ -1255,8 +1254,7 @@ write_archive (bfd *iarch) - bfd *contents_head = iarch->archive_next; - int ofd = -1; - -- old_name = (char *) xmalloc (strlen (bfd_get_filename (iarch)) + 1); -- strcpy (old_name, bfd_get_filename (iarch)); -+ old_name = xstrdup (bfd_get_filename (iarch)); - new_name = make_tempname (old_name, &ofd); - - if (new_name == NULL) -diff --git a/binutils/arsup.c b/binutils/arsup.c -index 0a1f63f6456..f7ce8f0bc82 100644 ---- a/binutils/arsup.c -+++ b/binutils/arsup.c -@@ -42,6 +42,8 @@ extern int deterministic; - - static bfd *obfd; - static char *real_name; -+static char *temp_name; -+static int real_ofd; - static FILE *outfile; - - static void -@@ -149,27 +151,24 @@ maybequit (void) - void - ar_open (char *name, int t) - { -- char *tname; -- const char *bname = lbasename (name); -- real_name = name; -+ real_name = xstrdup (name); -+ temp_name = make_tempname (real_name, &real_ofd); - -- /* Prepend tmp- to the beginning, to avoid file-name clashes after -- truncation on filesystems with limited namespaces (DOS). */ -- if (asprintf (&tname, "%.*stmp-%s", (int) (bname - name), name, bname) == -1) -+ if (temp_name == NULL) - { -- fprintf (stderr, _("%s: Can't allocate memory for temp name (%s)\n"), -+ fprintf (stderr, _("%s: Can't open temporary file (%s)\n"), - program_name, strerror(errno)); - maybequit (); - return; - } - -- obfd = bfd_openw (tname, NULL); -+ obfd = bfd_fdopenw (temp_name, NULL, real_ofd); - - if (!obfd) - { - fprintf (stderr, - _("%s: Can't open output archive %s\n"), -- program_name, tname); -+ program_name, temp_name); - - maybequit (); - } -@@ -344,16 +343,30 @@ ar_save (void) - } - else - { -- char *ofilename = xstrdup (bfd_get_filename (obfd)); -+ struct stat target_stat; - - if (deterministic > 0) - obfd->flags |= BFD_DETERMINISTIC_OUTPUT; - - bfd_close (obfd); - -- smart_rename (ofilename, real_name, NULL); -+ if (stat (real_name, &target_stat) != 0) -+ { -+ /* The temp file created in ar_open has mode 0600 as per mkstemp. -+ Create the real empty output file here so smart_rename will -+ update the mode according to the process umask. */ -+ obfd = bfd_openw (real_name, NULL); -+ if (obfd != NULL) -+ { -+ bfd_set_format (obfd, bfd_archive); -+ bfd_close (obfd); -+ } -+ } -+ -+ smart_rename (temp_name, real_name, NULL); - obfd = 0; -- free (ofilename); -+ free (temp_name); -+ free (real_name); - } - } - -diff --git a/binutils/objcopy.c b/binutils/objcopy.c -index 07a872b5a80..73aa8bc2514 100644 ---- a/binutils/objcopy.c -+++ b/binutils/objcopy.c -@@ -20,7 +20,6 @@ - - #include "sysdep.h" - #include "bfd.h" --#include "libbfd.h" - #include "progress.h" - #include "getopt.h" - #include "libiberty.h" -diff --git a/binutils/rename.c b/binutils/rename.c -index f471b45fd3f..2ff092ee22b 100644 ---- a/binutils/rename.c -+++ b/binutils/rename.c -@@ -130,11 +130,11 @@ int - smart_rename (const char *from, const char *to, - struct stat *target_stat ATTRIBUTE_UNUSED) - { -- bfd_boolean exists; -- struct stat s; - int ret = 0; -+ struct stat to_stat; -+ bfd_boolean exists; - -- exists = lstat (to, &s) == 0; -+ exists = lstat (to, &to_stat) == 0; - - #if defined (_WIN32) && !defined (__CYGWIN32__) - /* Win32, unlike unix, will not erase `to' in `rename(from, to)' but --- -2.31.1 - diff --git a/meta/recipes-devtools/binutils/binutils/0003-CVE-2021-20197.patch b/meta/recipes-devtools/binutils/binutils/0003-CVE-2021-20197.patch deleted file mode 100644 index 082b28b29c..0000000000 --- a/meta/recipes-devtools/binutils/binutils/0003-CVE-2021-20197.patch +++ /dev/null @@ -1,171 +0,0 @@ -From 8b69e61d4be276bb862698aaafddc3e779d23c8f Mon Sep 17 00:00:00 2001 -From: Alan Modra <amodra@gmail.com> -Date: Tue, 23 Feb 2021 09:37:39 +1030 -Subject: [PATCH] PR27456, lstat in rename.c on MinGW - - PR 27456 - * rename.c: Tidy throughout. - (smart_rename): Always copy. Remove windows specific code. - -(cherry picked from commit cca8873dd5a6015d5557ea44bc1ea9c252435a29) - -Upstream-Status: Backport [https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;h=8b69e61d4be276bb862698aaafddc3e779d23c8f] -CVE: CVE-2021-20197 -Signed-off-by: Vinay Kumar <vinay.m.engg@gmail.com> ---- - binutils/rename.c | 111 ++++++++++++++------------------------------- - 2 files changed, 40 insertions(+), 76 deletions(-) - -diff --git a/binutils/rename.c b/binutils/rename.c -index 2ff092ee22b..72a9323d72c 100644 ---- a/binutils/rename.c -+++ b/binutils/rename.c -@@ -24,14 +24,9 @@ - - #ifdef HAVE_GOOD_UTIME_H - #include <utime.h> --#else /* ! HAVE_GOOD_UTIME_H */ --#ifdef HAVE_UTIMES -+#elif defined HAVE_UTIMES - #include <sys/time.h> --#endif /* HAVE_UTIMES */ --#endif /* ! HAVE_GOOD_UTIME_H */ -- --#if ! defined (_WIN32) || defined (__CYGWIN32__) --static int simple_copy (const char *, const char *); -+#endif - - /* The number of bytes to copy at once. */ - #define COPY_BUF 8192 -@@ -82,7 +77,6 @@ simple_copy (const char *from, const char *to) - } - return 0; - } --#endif /* __CYGWIN32__ or not _WIN32 */ - - /* Set the times of the file DESTINATION to be the same as those in - STATBUF. */ -@@ -91,87 +85,52 @@ void - set_times (const char *destination, const struct stat *statbuf) - { - int result; -- -- { - #ifdef HAVE_GOOD_UTIME_H -- struct utimbuf tb; -- -- tb.actime = statbuf->st_atime; -- tb.modtime = statbuf->st_mtime; -- result = utime (destination, &tb); --#else /* ! HAVE_GOOD_UTIME_H */ --#ifndef HAVE_UTIMES -- long tb[2]; -- -- tb[0] = statbuf->st_atime; -- tb[1] = statbuf->st_mtime; -- result = utime (destination, tb); --#else /* HAVE_UTIMES */ -- struct timeval tv[2]; -- -- tv[0].tv_sec = statbuf->st_atime; -- tv[0].tv_usec = 0; -- tv[1].tv_sec = statbuf->st_mtime; -- tv[1].tv_usec = 0; -- result = utimes (destination, tv); --#endif /* HAVE_UTIMES */ --#endif /* ! HAVE_GOOD_UTIME_H */ -- } -+ struct utimbuf tb; -+ -+ tb.actime = statbuf->st_atime; -+ tb.modtime = statbuf->st_mtime; -+ result = utime (destination, &tb); -+#elif defined HAVE_UTIMES -+ struct timeval tv[2]; -+ -+ tv[0].tv_sec = statbuf->st_atime; -+ tv[0].tv_usec = 0; -+ tv[1].tv_sec = statbuf->st_mtime; -+ tv[1].tv_usec = 0; -+ result = utimes (destination, tv); -+#else -+ long tb[2]; -+ -+ tb[0] = statbuf->st_atime; -+ tb[1] = statbuf->st_mtime; -+ result = utime (destination, tb); -+#endif - - if (result != 0) - non_fatal (_("%s: cannot set time: %s"), destination, strerror (errno)); - } - --/* Rename FROM to TO, copying if TO exists. TARGET_STAT has the file status -- that, if non-NULL, is used to fix up timestamps after rename. Return 0 if -- ok, -1 if error. */ -+/* Copy FROM to TO. TARGET_STAT has the file status that, if non-NULL, -+ is used to fix up timestamps. Return 0 if ok, -1 if error. -+ At one time this function renamed files, but file permissions are -+ tricky to update given the number of different schemes used by -+ various systems. So now we just copy. */ - - int - smart_rename (const char *from, const char *to, -- struct stat *target_stat ATTRIBUTE_UNUSED) -+ struct stat *target_stat) - { -- int ret = 0; -- struct stat to_stat; -- bfd_boolean exists; -- -- exists = lstat (to, &to_stat) == 0; -- --#if defined (_WIN32) && !defined (__CYGWIN32__) -- /* Win32, unlike unix, will not erase `to' in `rename(from, to)' but -- fail instead. Also, chown is not present. */ -- -- if (exists) -- remove (to); -+ int ret; - -- ret = rename (from, to); -+ ret = simple_copy (from, to); - if (ret != 0) -- { -- /* We have to clean up here. */ -- non_fatal (_("unable to rename '%s'; reason: %s"), to, strerror (errno)); -- unlink (from); -- } --#else -- /* Avoid a full copy and use rename if TO does not exist. */ -- if (!exists) -- { -- if ((ret = rename (from, to)) != 0) -- { -- /* We have to clean up here. */ -- non_fatal (_("unable to rename '%s'; reason: %s"), to, strerror (errno)); -- unlink (from); -- } -- } -- else -- { -- ret = simple_copy (from, to); -- if (ret != 0) -- non_fatal (_("unable to copy file '%s'; reason: %s"), to, strerror (errno)); -+ non_fatal (_("unable to copy file '%s'; reason: %s"), -+ to, strerror (errno)); - -- if (target_stat != NULL) -- set_times (to, target_stat); -- unlink (from); -- } --#endif /* _WIN32 && !__CYGWIN32__ */ -+ if (target_stat != NULL) -+ set_times (to, target_stat); -+ unlink (from); - - return ret; - } --- -2.31.1 - diff --git a/meta/recipes-devtools/binutils/binutils/0017-CVE-2021-3530.patch b/meta/recipes-devtools/binutils/binutils/0017-CVE-2021-3530.patch new file mode 100644 index 0000000000..fa10a6c99a --- /dev/null +++ b/meta/recipes-devtools/binutils/binutils/0017-CVE-2021-3530.patch @@ -0,0 +1,102 @@ +From 25162c795b1a2becf936bb3581d86a307ea491eb Mon Sep 17 00:00:00 2001 +From: Nick Clifton <nickc@redhat.com> +Date: Thu, 15 Jul 2021 16:51:56 +0100 +Subject: [PATCH] Fix a stack exhaustion problem in the Rust demangling code in + the libiberty library. + + PR 99935 + * rust-demangle.c: Add recursion limit. + +CVE: CVE-2021-3530 +Upstream-Status: Backport[https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=25162c795b1a2becf936bb3581d86a307ea491eb] +Signed-off-by: Pgowda <pgowda.cve@gmail.com> + +--- + libiberty/ChangeLog | 5 +++++ + libiberty/rust-demangle.c | 31 +++++++++++++++++++++++++------ + 2 files changed, 30 insertions(+), 6 deletions(-) + +diff --git a/libiberty/ChangeLog b/libiberty/ChangeLog +index bc1b35b97c4..8e39fd28eba 100644 +--- a/libiberty/ChangeLog ++++ b/libiberty/ChangeLog +@@ -1,3 +1,8 @@ ++2021-07-15 Nick Clifton <nickc@redhat.com> ++ ++ PR 99935 ++ * rust-demangle.c: Add recursion limit. ++ + 2021-01-04 Martin Liska <mliska@suse.cz> + + * strverscmp.c: Convert to utf8 from iso8859. +diff --git a/libiberty/rust-demangle.c b/libiberty/rust-demangle.c +index 449941b56dc..df09b7b8fdd 100644 +--- a/libiberty/rust-demangle.c ++++ b/libiberty/rust-demangle.c +@@ -74,6 +74,12 @@ struct rust_demangler + /* Rust mangling version, with legacy mangling being -1. */ + int version; + ++ /* Recursion depth. */ ++ uint recursion; ++ /* Maximum number of times demangle_path may be called recursively. */ ++#define RUST_MAX_RECURSION_COUNT 1024 ++#define RUST_NO_RECURSION_LIMIT ((uint) -1) ++ + uint64_t bound_lifetime_depth; + }; + +@@ -671,6 +677,15 @@ demangle_path (struct rust_demangler *rd + if (rdm->errored) + return; + ++ if (rdm->recursion != RUST_NO_RECURSION_LIMIT) ++ { ++ ++ rdm->recursion; ++ if (rdm->recursion > RUST_MAX_RECURSION_COUNT) ++ /* FIXME: There ought to be a way to report ++ that the recursion limit has been reached. */ ++ goto fail_return; ++ } ++ + switch (tag = next (rdm)) + { + case 'C': +@@ -688,10 +703,7 @@ demangle_path (struct rust_demangler *rd + case 'N': + ns = next (rdm); + if (!ISLOWER (ns) && !ISUPPER (ns)) +- { +- rdm->errored = 1; +- return; +- } ++ goto fail_return; + + demangle_path (rdm, in_value); + +@@ -776,9 +788,15 @@ demangle_path (struct rust_demangler *rd + } + break; + default: +- rdm->errored = 1; +- return; ++ goto fail_return; + } ++ goto pass_return; ++ ++ fail_return: ++ rdm->errored = 1; ++ pass_return: ++ if (rdm->recursion != RUST_NO_RECURSION_LIMIT) ++ -- rdm->recursion; + } + + static void +@@ -1317,6 +1335,7 @@ rust_demangle_callback (const char *mang + rdm.skipping_printing = 0; + rdm.verbose = (options & DMGL_VERBOSE) != 0; + rdm.version = 0; ++ rdm.recursion = (options & DMGL_NO_RECURSE_LIMIT) ? RUST_NO_RECURSION_LIMIT : 0; + rdm.bound_lifetime_depth = 0; + + /* Rust symbols always start with _R (v0) or _ZN (legacy). */ diff --git a/meta/recipes-devtools/binutils/binutils/0018-CVE-2021-3530.patch b/meta/recipes-devtools/binutils/binutils/0018-CVE-2021-3530.patch new file mode 100644 index 0000000000..e569a6bc47 --- /dev/null +++ b/meta/recipes-devtools/binutils/binutils/0018-CVE-2021-3530.patch @@ -0,0 +1,64 @@ +From 999566402e3d7c69032bbf47e28b44fc0926fe62 Mon Sep 17 00:00:00 2001 +From: Christopher Wellons <wellons@nullprogram.com> +Date: Sun, 18 Jul 2021 16:57:19 -0400 +Subject: [PATCH] Change "uint" to "unsigned" + +This fixes a defect introduced in 25162c795. The "uint" type has not +been explicitly defined here on mingw, causing compilation to fail. + +On linux we have this in /usr/include/sys/types.h + +/* Old compatibility names for C types. */ +typedef unsigned long int ulong; +typedef unsigned short int ushort; +typedef unsigned int uint; + +So it's easy to see how such bugs can creep in. + + * rust-demangle.c (struct rust_demangler): Change type of + "recursion" to unsigned. + (RUST_NO_RECURSION_LIMIT): Similarly in cast. + +CVE: CVE-2021-3530 +Upstream-Status: Backport[https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=999566402e3] +Signed-off-by: Pgowda <pgowda.cve@gmail.com> + +--- + libiberty/ChangeLog | 6 ++++++ + libiberty/rust-demangle.c | 4 ++-- + 2 files changed, 8 insertions(+), 2 deletions(-) + +diff --git a/libiberty/ChangeLog b/libiberty/ChangeLog +index 8e39fd28eba..3f749455f05 100644 +--- a/libiberty/ChangeLog ++++ b/libiberty/ChangeLog +@@ -1,3 +1,9 @@ ++2021-07-19 Christopher Wellons <wellons@nullprogram.com> ++ ++ * rust-demangle.c (struct rust_demangler): Change type of ++ "recursion" to unsigned. ++ (RUST_NO_RECURSION_LIMIT): Similarly in cast. ++ + 2021-07-15 Nick Clifton <nickc@redhat.com> + + PR 99935 +diff --git a/libiberty/rust-demangle.c b/libiberty/rust-demangle.c +index df09b7b8fdd..ac1eb8eb02c 100644 +--- a/libiberty/rust-demangle.c ++++ b/libiberty/rust-demangle.c +@@ -75,10 +75,10 @@ struct rust_demangler + int version; + + /* Recursion depth. */ +- uint recursion; ++ unsigned recursion; + /* Maximum number of times demangle_path may be called recursively. */ + #define RUST_MAX_RECURSION_COUNT 1024 +-#define RUST_NO_RECURSION_LIMIT ((uint) -1) ++#define RUST_NO_RECURSION_LIMIT ((unsigned) -1) + + uint64_t bound_lifetime_depth; + }; +-- +2.27.0 + |