All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Mittal, Anuj" <anuj.mittal@intel.com>
To: "openembedded-core@lists.openembedded.org"
	<openembedded-core@lists.openembedded.org>,
	"pgowda.cve@gmail.com" <pgowda.cve@gmail.com>
Cc: "richard.purdie@linuxfoundation.org"
	<richard.purdie@linuxfoundation.org>,
	"rwmacleod@gmail.com" <rwmacleod@gmail.com>,
	"umesh.kalappa0@gmail.com" <umesh.kalappa0@gmail.com>
Subject: Re: [hardknott][PATCH] glibc: Backport fix for CVE-2021-43396
Date: Mon, 10 Jan 2022 04:45:14 +0000	[thread overview]
Message-ID: <1e4506fb15fee3db97d0549479ed7c15bbb72f91.camel@intel.com> (raw)
In-Reply-To: <20211202073845.150820-1-pgowda.cve@gmail.com>

On Wed, 2021-12-01 at 23:38 -0800, pgowda wrote:
> Backport the fix for CVE-2021-43396. It is disputed that this is a
> security issue.
> 
> (From OE-Core rev: e8de9b01c6b305b2498c5f942397a49ae2af0cde)
> 
> Signed-off-by: pgowda <pgowda.cve@gmail.com>
> ---
>  .../glibc/glibc/0031-CVE-2021-43396.patch     | 188
> ++++++++++++++++++
>  meta/recipes-core/glibc/glibc_2.33.bb         |   1 +
>  2 files changed, 189 insertions(+)
>  create mode 100644 meta/recipes-core/glibc/glibc/0031-CVE-2021-
> 43396.patch

This doesn't apply cleanly:

WARNING: glibc-2.33-r0 do_patch: Fuzz detected:

Applying patch 0031-CVE-2021-43396.patch
patching file iconvdata/Makefile
Hunk #3 succeeded at 325 with fuzz 2 (offset -3 lines).
patching file iconvdata/bug-iconv15.c
patching file iconvdata/iso-2022-jp-3.c
Hunk #1 succeeded at 1 with fuzz 2.
Hunk #2 succeeded at 82 (offset 2 lines).


The context lines in the patches can be updated with devtool:

    devtool modify glibc
    devtool finish --force-patch-refresh glibc <layer_path>

Don't forget to review changes done by devtool!

Thanks,

Anuj

> 
> diff --git a/meta/recipes-core/glibc/glibc/0031-CVE-2021-43396.patch
> b/meta/recipes-core/glibc/glibc/0031-CVE-2021-43396.patch
> new file mode 100644
> index 0000000000..179b0c559d
> --- /dev/null
> +++ b/meta/recipes-core/glibc/glibc/0031-CVE-2021-43396.patch
> @@ -0,0 +1,188 @@
> +From ff012870b2c02a62598c04daa1e54632e020fd7d Mon Sep 17 00:00:00
> 2001
> +From: Nikita Popov <npv1310@gmail.com>
> +Date: Tue, 2 Nov 2021 13:21:42 +0500
> +Subject: [PATCH] gconv: Do not emit spurious NUL character in ISO-
> 2022-JP-3
> + (bug 28524)
> +
> +Bugfix 27256 has introduced another issue:
> +In conversion from ISO-2022-JP-3 encoding, it is possible
> +to force iconv to emit extra NUL character on internal state reset.
> +To do this, it is sufficient to feed iconv with escape sequence
> +which switches active character set.
> +The simplified check 'data->__statep->__count != ASCII_set'
> +introduced by the aforementioned bugfix picks that case and
> +behaves as if '\0' character has been queued thus emitting it.
> +
> +To eliminate this issue, these steps are taken:
> +* Restore original condition
> +'(data->__statep->__count & ~7) != ASCII_set'.
> +It is necessary since bits 0-2 may contain
> +number of buffered input characters.
> +* Check that queued character is not NUL.
> +Similar step is taken for main conversion loop.
> +
> +Bundled test case follows following logic:
> +* Try to convert ISO-2022-JP-3 escape sequence
> +switching active character set
> +* Reset internal state by providing NULL as input buffer
> +* Ensure that nothing has been converted.
> +
> +Signed-off-by: Nikita Popov <npv1310@gmail.com>
> +
> +CVE: CVE-2021-43396
> +Upstream-Status: Backport [ff012870b2c02a62598c04daa1e54632e020fd7d]
> +---
> + iconvdata/Makefile        |  5 +++-
> + iconvdata/bug-iconv15.c   | 60
> +++++++++++++++++++++++++++++++++++++++
> + iconvdata/iso-2022-jp-3.c | 28 ++++++++++++------
> + 3 files changed, 84 insertions(+), 9 deletions(-)
> + create mode 100644 iconvdata/bug-iconv15.c
> +
> +diff --git a/iconvdata/Makefile b/iconvdata/Makefile
> +index c216f959df..d5507a048c 100644
> +--- a/iconvdata/Makefile
> ++++ b/iconvdata/Makefile
> +@@ -1,4 +1,5 @@
> + # Copyright (C) 1997-2021 Free Software Foundation, Inc.
> ++# Copyright (C) The GNU Toolchain Authors.
> + # This file is part of the GNU C Library.
> + 
> + # The GNU C Library is free software; you can redistribute it
> and/or
> +@@ -74,7 +75,7 @@ ifeq (yes,$(build-shared))
> + tests = bug-iconv1 bug-iconv2 tst-loading tst-e2big tst-iconv4 bug-
> iconv4 \
> +       tst-iconv6 bug-iconv5 bug-iconv6 tst-iconv7 bug-iconv8 bug-
> iconv9 \
> +       bug-iconv10 bug-iconv11 bug-iconv12 tst-iconv-big5-hkscs-to-
> 2ucs4 \
> +-      bug-iconv13 bug-iconv14
> ++      bug-iconv13 bug-iconv14 bug-iconv15
> + ifeq ($(have-thread-library),yes)
> + tests += bug-iconv3
> + endif
> +@@ -327,6 +328,8 @@ $(objpfx)bug-iconv12.out: $(addprefix $(objpfx),
> $(gconv-modules)) \
> +                         $(addprefix $(objpfx),$(modules.so))
> + $(objpfx)bug-iconv14.out: $(addprefix $(objpfx), $(gconv-modules))
> \
> +                         $(addprefix $(objpfx),$(modules.so))
> ++$(objpfx)bug-iconv15.out: $(addprefix $(objpfx), $(gconv-modules))
> \
> ++                        $(addprefix $(objpfx),$(modules.so))
> + 
> + $(objpfx)iconv-test.out: run-iconv-test.sh \
> +                        $(addprefix $(objpfx), $(gconv-modules)) \
> +diff --git a/iconvdata/bug-iconv15.c b/iconvdata/bug-iconv15.c
> +new file mode 100644
> +index 0000000000..cc04bd0313
> +--- /dev/null
> ++++ b/iconvdata/bug-iconv15.c
> +@@ -0,0 +1,60 @@
> ++/* Bug 28524: Conversion from ISO-2022-JP-3 with iconv
> ++   may emit spurious NUL character on state reset.
> ++   Copyright (C) The GNU Toolchain Authors.
> ++   This file is part of the GNU C Library.
> ++
> ++   The GNU C Library is free software; you can redistribute it
> and/or
> ++   modify it under the terms of the GNU Lesser General Public
> ++   License as published by the Free Software Foundation; either
> ++   version 2.1 of the License, or (at your option) any later
> version.
> ++
> ++   The GNU C Library is distributed in the hope that it will be
> useful,
> ++   but WITHOUT ANY WARRANTY; without even the implied warranty of
> ++   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> GNU
> ++   Lesser General Public License for more details.
> ++
> ++   You should have received a copy of the GNU Lesser General Public
> ++   License along with the GNU C Library; if not, see
> ++   <https://www.gnu.org/licenses/>.  */
> ++
> ++#include <stddef.h>
> ++#include <iconv.h>
> ++#include <support/check.h>
> ++
> ++static int
> ++do_test (void)
> ++{
> ++  char in[] = "\x1b(I";
> ++  char *inbuf = in;
> ++  size_t inleft = sizeof (in) - 1;
> ++  char out[1];
> ++  char *outbuf = out;
> ++  size_t outleft = sizeof (out);
> ++  iconv_t cd;
> ++
> ++  cd = iconv_open ("UTF8", "ISO-2022-JP-3");
> ++  TEST_VERIFY_EXIT (cd != (iconv_t) -1);
> ++
> ++  /* First call to iconv should alter internal state.
> ++     Now, JISX0201_Kana_set is selected and
> ++     state value != ASCII_set.  */
> ++  TEST_VERIFY (iconv (cd, &inbuf, &inleft, &outbuf, &outleft) !=
> (size_t) -1);
> ++
> ++  /* No bytes should have been added to
> ++     the output buffer at this point.  */
> ++  TEST_VERIFY (outbuf == out);
> ++  TEST_VERIFY (outleft == sizeof (out));
> ++
> ++  /* Second call shall emit spurious NUL character in unpatched
> glibc.  */
> ++  TEST_VERIFY (iconv (cd, NULL, NULL, &outbuf, &outleft) !=
> (size_t) -1);
> ++
> ++  /* No characters are expected to be produced.  */
> ++  TEST_VERIFY (outbuf == out);
> ++  TEST_VERIFY (outleft == sizeof (out));
> ++
> ++  TEST_VERIFY_EXIT (iconv_close (cd) != -1);
> ++
> ++  return 0;
> ++}
> ++
> ++#include <support/test-driver.c>
> +diff --git a/iconvdata/iso-2022-jp-3.c b/iconvdata/iso-2022-jp-3.c
> +index 70b28ace7f..d724126545 100644
> +--- a/iconvdata/iso-2022-jp-3.c
> ++++ b/iconvdata/iso-2022-jp-3.c
> +@@ -1,5 +1,6 @@
> + /* Conversion module for ISO-2022-JP-3.
> +    Copyright (C) 1998-2021 Free Software Foundation, Inc.
> ++   Copyright (C) The GNU Toolchain Authors.
> +    This file is part of the GNU C Library.
> + 
> +    The GNU C Library is free software; you can redistribute it
> and/or
> +@@ -79,20 +80,31 @@ enum
> +    the output state to the initial state.  This has to be done
> during the
> +    flushing.  */
> + #define EMIT_SHIFT_TO_INIT \
> +-  if (data->__statep->__count != ASCII_set)                        
> \
> ++  if ((data->__statep->__count & ~7) !=
> ASCII_set)                          \
> +    
> {                                                                    
>    \
> +       if
> (FROM_DIRECTION)                                                   \
> +       {                                                            
>          \
> +-        if (__glibc_likely (outbuf + 4 <=
> outend))                          \
> ++        uint32_t ch = data->__statep->__count >>
> 6;                         \
> ++                                                                   
>          \
> ++        if (__glibc_unlikely (ch !=
> 0))                                     \
> +          
> {                                                                 \
> +-            /* Write out the last character. 
> */                            \
> +-            *((uint32_t *) outbuf) = data->__statep->__count >>
> 6;          \
> +-            outbuf += sizeof
> (uint32_t);                                    \
> +-            data->__statep->__count =
> ASCII_set;                      \
> ++            if (__glibc_likely (outbuf + 4 <=
> outend))                      \
> ++              {                                                    
>          \
> ++                /* Write out the last character. 
> */                        \
> ++                put32u (outbuf,
> ch);                                        \
> ++                outbuf +=
> 4;                                                \
> ++                data->__statep->__count &=
> 7;                               \
> ++                data->__statep->__count |=
> ASCII_set;                       \
> ++              }                                                    
>          \
> ++           
> else                                                            \
> ++              /* We don't have enough room in the output buffer. 
> */        \
> ++              status =
> __GCONV_FULL_OUTPUT;                                 \
> +          
> }                                                                 \
> +        
> else                                                                \
> +-          /* We don't have enough room in the output buffer. 
> */            \
> +-          status =
> __GCONV_FULL_OUTPUT;                                     \
> ++         
> {                                                                 \
> ++            data->__statep->__count &=
> 7;                                   \
> ++            data->__statep->__count |=
> ASCII_set;                           \
> ++         
> }                                                                 \
> +       }                                                            
>          \
> +      
> else                                                                 
> \
> +       {                                                            
>          \
> +-- 
> +2.27.0
> +
> diff --git a/meta/recipes-core/glibc/glibc_2.33.bb b/meta/recipes-
> core/glibc/glibc_2.33.bb
> index ad5e2b8eb1..38b8a8b065 100644
> --- a/meta/recipes-core/glibc/glibc_2.33.bb
> +++ b/meta/recipes-core/glibc/glibc_2.33.bb
> @@ -56,6 +56,7 @@ SRC_URI = 
> "${GLIBC_GIT_URI};branch=${SRCBRANCH};name=glibc \
>            
> file://0028-readlib-Add-OECORE_KNOWN_INTERPRETER_NAMES-to-known-.patch
>  \
>            
> file://0029-wordsize.h-Unify-the-header-between-arm-and-aarch64.patch
>  \
>            
> file://0030-powerpc-Do-not-ask-compiler-for-finding-arch.patch \
> +           file://0031-CVE-2021-43396.patch \
>             "
>  S = "${WORKDIR}/git"
>  B = "${WORKDIR}/build-${TARGET_SYS}"


      parent reply	other threads:[~2022-01-10  4:45 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-12-02  7:38 [hardknott][PATCH] glibc: Backport fix for CVE-2021-43396 pgowda
2022-01-06 10:40 ` pgowda cve
2022-01-10  4:45 ` Mittal, Anuj [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1e4506fb15fee3db97d0549479ed7c15bbb72f91.camel@intel.com \
    --to=anuj.mittal@intel.com \
    --cc=openembedded-core@lists.openembedded.org \
    --cc=pgowda.cve@gmail.com \
    --cc=richard.purdie@linuxfoundation.org \
    --cc=rwmacleod@gmail.com \
    --cc=umesh.kalappa0@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.