GNU bug report logs - #18784
Coultdn't compile emacs-24.4

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 18784 in the body.
You can then email your comments to 18784 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Gangræna Gorgeous <trupanka <at> gmail.com>
To: bug-gnu-emacs <at> gnu.org
Subject: Coultdn't compile emacs-24.4
Date: Tue, 21 Oct 2014 15:52:40 +0400

[Message part 1 (text/plain, inline)]

I've got segfault trying to compile emacs-24.4.

dmesg:
[ 1807.828767] grsec: denied resource overstep by requesting 4294967295 for
RLIMIT_NOFILE against limit 1048576 for
/var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/conftest[conftest:8947]
uid/euid:0/0 gid/egid:0/0, parent
/var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/configure[configure:8944]
uid/euid:0/0 gid/egid:0/0
[ 1807.828805] grsec: denied resource overstep by requesting 1048576 for
RLIMIT_NOFILE against limit 1048576 for
/var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/conftest[conftest:8947]
uid/euid:0/0 gid/egid:0/0, parent
/var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/configure[configure:8944]
uid/euid:0/0 gid/egid:0/0
[ 1817.329338] grsec: denied resource overstep by requesting 1048576 for
RLIMIT_NOFILE against limit 1048576 for
/var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/conftest[conftest:9974]
uid/euid:0/0 gid/egid:0/0, parent
/var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/configure[configure:9973]
uid/euid:0/0 gid/egid:0/0
[ 1850.050464] temacs[11504]: segfault at 47e1c0 ip 000003fff51346d0 sp
000003ffffff8478 error 4 in libc-2.19.so[3fff509b000+1af000]

[Message part 2 (text/html, inline)]

[build.log (text/x-log, attachment)]

Message #8 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Gangræna Gorgeous <trupanka <at> gmail.com>
Cc: 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Tue, 21 Oct 2014 18:16:00 +0300

> Date: Tue, 21 Oct 2014 15:52:40 +0400
> From: Gangræna Gorgeous <trupanka <at> gmail.com>
> 
> I've got segfault trying to compile emacs-24.4.
> 
> dmesg:
> [ 1807.828767] grsec: denied resource overstep by requesting 4294967295 for
> RLIMIT_NOFILE against limit 1048576 for
> /var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/conftest[conftest:8947]
> uid/euid:0/0 gid/egid:0/0, parent
> /var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/configure
> [configure:8944] uid/euid:0/0 gid/egid:0/0
> [ 1807.828805] grsec: denied resource overstep by requesting 1048576 for
> RLIMIT_NOFILE against limit 1048576 for
> /var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/conftest[conftest:8947]
> uid/euid:0/0 gid/egid:0/0, parent
> /var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/configure
> [configure:8944] uid/euid:0/0 gid/egid:0/0
> [ 1817.329338] grsec: denied resource overstep by requesting 1048576 for
> RLIMIT_NOFILE against limit 1048576 for
> /var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/conftest[conftest:9974]
> uid/euid:0/0 gid/egid:0/0, parent
> /var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/configure
> [configure:9973] uid/euid:0/0 gid/egid:0/0
> [ 1850.050464] temacs[11504]: segfault at 47e1c0 ip 000003fff51346d0 sp
> 000003ffffff8478 error 4 in libc-2.19.so[3fff509b000+1af000]

Sounds like some locale problem with the system resources you are
allowed to use.

Message #11 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: David Engster <deng <at> randomsample.de>
To: Eli Zaretskii <eliz <at> gnu.org>
Cc: Gangræna Gorgeous <trupanka <at> gmail.com>,
 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Tue, 21 Oct 2014 17:49:42 +0200

Eli Zaretskii writes:
>> Date: Tue, 21 Oct 2014 15:52:40 +0400
>> From: Gangræna Gorgeous <trupanka <at> gmail.com>
>> 
>
>> I've got segfault trying to compile emacs-24.4.
>> 
>> dmesg:
>> [ 1807.828767] grsec: denied resource overstep by requesting 4294967295 for
>> RLIMIT_NOFILE against limit 1048576 for
>> /var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/conftest[conftest:8947]
>> uid/euid:0/0 gid/egid:0/0, parent
>> /var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/configure
>> [configure:8944] uid/euid:0/0 gid/egid:0/0
>> [ 1807.828805] grsec: denied resource overstep by requesting 1048576 for
>> RLIMIT_NOFILE against limit 1048576 for
>> /var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/conftest[conftest:8947]
>> uid/euid:0/0 gid/egid:0/0, parent
>> /var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/configure
>> [configure:8944] uid/euid:0/0 gid/egid:0/0
>> [ 1817.329338] grsec: denied resource overstep by requesting 1048576 for
>> RLIMIT_NOFILE against limit 1048576 for
>> /var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/conftest[conftest:9974]
>> uid/euid:0/0 gid/egid:0/0, parent
>> /var/tmp/portage/app-editors/emacs-24.4/work/emacs-24.4/configure
>> [configure:9973] uid/euid:0/0 gid/egid:0/0
>> [ 1850.050464] temacs[11504]: segfault at 47e1c0 ip 000003fff51346d0 sp
>> 000003ffffff8478 error 4 in libc-2.19.so[3fff509b000+1af000]
>
> Sounds like some locale problem with the system resources you are
> allowed to use.

From the looks of it the OP seems to use Gentoo with PaX patches, so
this could be related to #11398.

-David

Message #16 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Glenn Morris <rgm <at> gnu.org>
To: Gangræna Gorgeous <trupanka <at> gmail.com>
Cc: Ulrich Mueller <ulm <at> gentoo.org>, 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Fri, 13 Mar 2015 15:54:32 -0400

Sorry for the lack of response.
You might have more luck reporting this to Gentoo.

Perhaps still more is needed beyond what was installed for
http://debbugs.gnu.org/11398 and http://debbugs.gnu.org/16343,
but I have no idea what.

Message #19 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Ulrich Mueller <ulm <at> gentoo.org>
To: Glenn Morris <rgm <at> gnu.org>
Cc: Gangræna Gorgeous <trupanka <at> gmail.com>,
 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Fri, 13 Mar 2015 21:43:25 +0100

>>>>> On Fri, 13 Mar 2015, Glenn Morris wrote:

> Sorry for the lack of response. You might have more luck reporting
> this to Gentoo.

> Perhaps still more is needed beyond what was installed for
> http://debbugs.gnu.org/11398 and http://debbugs.gnu.org/16343, but I
> have no idea what.

This doesn't look like a Linux PaX kernel issue to me, but rather like
overly aggressive CFLAGS and LDFLAGS (see build.log):
 * strip-flags: CFLAGS: changed '-O2 -pipe -m64 -march=amdfam10 -m3dnow -mmmx -msse -msse2 -msse3 -msse4a -mcx16 -msahf -mfpmath=sse -mabm -m128bit-long-double -fPIE -U_FORTIFY_SOURCE -D_FORTIFY_SOURCE=2' to '-O2 -pipe -m64 -march=amdfam10 -U_FORTIFY_SOURCE -D_FORTIFY_SOURCE=2'
 * strip-flags: LDFLAGS: changed '-pie -m64 -Wl,-O2 -Wl,--as-needed -s -Wl,-z,now -Wl,-z,relro -Wl,-z,noexecstack -Wl,--hash-style=gnu' to '-m64 -Wl,-O2 -Wl,--as-needed -Wl,-z,now -Wl,-z,relro -Wl,-z,noexecstack -Wl,--hash-style=gnu'

But yes, if this is still an issue, please report it to the Gentoo bug
tracker: http://bugs.gentoo.org/

Message #22 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Ulrich Mueller <ulm <at> gentoo.org>
To: 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Fri, 13 Mar 2015 22:09:52 +0100

Thinking about it, this bug is most likely a duplicate of
http://debbugs.gnu.org/18780.

Gentoo explicitly adds -nopie to flags as a workaround since
app-editors/emacs-24.4-r1 (whereas the reporter has 24.4-r0 where the
problem is still present).

Message #29 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Stefan Monnier <monnier <at> iro.umontreal.ca>
To: Ulrich Mueller <ulm <at> gentoo.org>
Cc: 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Fri, 13 Mar 2015 19:25:35 -0400

forcemerge 18780 18784

> Thinking about it, this bug is most likely a duplicate of
> http://debbugs.gnu.org/18780.

> Gentoo explicitly adds -nopie to flags as a workaround since
> app-editors/emacs-24.4-r1 (whereas the reporter has 24.4-r0 where the
> problem is still present).

Indeed, looks right.  And 18780's diagnostic looks right as well.
Could someone look at fixing our configure script accordingly?


        Stefan

Message #34 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Ulrich Mueller <ulm <at> gentoo.org>
To: Stefan Monnier <monnier <at> iro.umontreal.ca>
Cc: 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Mon, 16 Mar 2015 10:16:37 +0100

>>>>> On Fri, 13 Mar 2015, Stefan Monnier wrote:

> Indeed, looks right. And 18780's diagnostic looks right as well.
> Could someone look at fixing our configure script accordingly?

Find my attempt of a patch below. Please review.


From 30678f8b5cba82b4ca9841fd1c8760b48cf12b59 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Ulrich=20M=C3=BCller?= <ulm <at> gentoo.org>
Date: Mon, 16 Mar 2015 10:02:51 +0100
Subject: [PATCH] configure.ac: Add -nopie to LD_SWITCH_SYSTEM_TEMACS.

* configure.ac (LD_SWITCH_SYSTEM_TEMACS): Add -nopie option,
in order to avoid segfaults in temacs.  (Bug#18780)
---
 ChangeLog    | 5 +++++
 configure.ac | 7 +++++--
 2 files changed, 10 insertions(+), 2 deletions(-)

diff --git a/ChangeLog b/ChangeLog
index 4f6523e..fd99ff9 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,8 @@
+2015-03-16  Ulrich Müller  <ulm <at> gentoo.org>
+
+	* configure.ac (LD_SWITCH_SYSTEM_TEMACS): Add -nopie option,
+	in order to avoid segfaults in temacs.  (Bug#18780)
+
 2015-03-02  Robert Pluim  <rpluim <at> gmail.com>  (tiny change)
 
 	* configure.ac: Error out if with-file-notification=w32 is
diff --git a/configure.ac b/configure.ac
index d65494a..0c0c5b6 100644
--- a/configure.ac
+++ b/configure.ac
@@ -4989,11 +4989,14 @@ case "$opsys" in
    esac
    ;;
 
-  openbsd) LD_SWITCH_SYSTEM_TEMACS='-nopie' ;;
-
   *) LD_SWITCH_SYSTEM_TEMACS= ;;
 esac
 
+if test "$GCC" = yes; then
+  # Disable PIE to avoid segfaults in temacs (bug#18780)
+  LD_SWITCH_SYSTEM_TEMACS="$LD_SWITCH_SYSTEM_TEMACS -nopie"
+fi
+
 if test x$ac_enable_profiling != x ; then
   case $opsys in
     *freebsd | gnu-linux) ;;
-- 
2.3.2

Message #37 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Stefan Monnier <monnier <at> iro.umontreal.ca>
To: Ulrich Mueller <ulm <at> gentoo.org>
Cc: 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Mon, 16 Mar 2015 09:09:52 -0400

> +if test "$GCC" = yes; then
> +  # Disable PIE to avoid segfaults in temacs (bug#18780)
> +  LD_SWITCH_SYSTEM_TEMACS="$LD_SWITCH_SYSTEM_TEMACS -nopie"
> +fi

Hmm... is `-nopie' a linker-option or a compiler-option?
IOW what about systems which use gcc to compile but something else to
link or vice-versa?


        Stefan

Message #40 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Stefan Monnier <monnier <at> iro.umontreal.ca>
Cc: ulm <at> gentoo.org, 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Mon, 16 Mar 2015 18:18:28 +0200

> From: Stefan Monnier <monnier <at> iro.umontreal.ca>
> Date: Mon, 16 Mar 2015 09:09:52 -0400
> Cc: 18784 <at> debbugs.gnu.org
> 
> > +if test "$GCC" = yes; then
> > +  # Disable PIE to avoid segfaults in temacs (bug#18780)
> > +  LD_SWITCH_SYSTEM_TEMACS="$LD_SWITCH_SYSTEM_TEMACS -nopie"
> > +fi
> 
> Hmm... is `-nopie' a linker-option or a compiler-option?

It's a GCC option, but at least my GCC doesn't recognize -nopie.  It
does recognize -fno-pie.

Anyway, please don't make that effective for every supported system;
e.g., I think it's wrong to do that for Windows, which always produces
PIE code.

Message #43 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Ulrich Mueller <ulm <at> gentoo.org>
To: Stefan Monnier <monnier <at> iro.umontreal.ca>
Cc: 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Mon, 16 Mar 2015 17:26:32 +0100

>>>>> On Mon, 16 Mar 2015, Stefan Monnier wrote:

> Hmm... is `-nopie' a linker-option or a compiler-option?
> IOW what about systems which use gcc to compile but something else
> to link or vice-versa?

Right. Also it seems that not all configurations of gcc will recognise
the -nopie option, so we must laboriously test for it. Updated patch
below.


From b3d5b10c10d7062b0da5c8316e787ae554f5c81a Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Ulrich=20M=C3=BCller?= <ulm <at> gentoo.org>
Date: Mon, 16 Mar 2015 17:09:54 +0100
Subject: [PATCH] configure.ac: Add -nopie to LD_SWITCH_SYSTEM_TEMACS.

* configure.ac (LD_SWITCH_SYSTEM_TEMACS): Add -nopie option if it
is supported, in order to avoid segfaults in temacs.  (Bug#18780)
---
 ChangeLog    |  5 +++++
 configure.ac | 15 +++++++++++++--
 2 files changed, 18 insertions(+), 2 deletions(-)

diff --git a/ChangeLog b/ChangeLog
index 4f6523e..308b889 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,8 @@
+2015-03-16  Ulrich Müller  <ulm <at> gentoo.org>
+
+	* configure.ac (LD_SWITCH_SYSTEM_TEMACS): Add -nopie option if it
+	is supported, in order to avoid segfaults in temacs.  (Bug#18780)
+
 2015-03-02  Robert Pluim  <rpluim <at> gmail.com>  (tiny change)
 
 	* configure.ac: Error out if with-file-notification=w32 is
diff --git a/configure.ac b/configure.ac
index d65494a..40ea5cb 100644
--- a/configure.ac
+++ b/configure.ac
@@ -4989,11 +4989,22 @@ case "$opsys" in
    esac
    ;;
 
-  openbsd) LD_SWITCH_SYSTEM_TEMACS='-nopie' ;;
-
   *) LD_SWITCH_SYSTEM_TEMACS= ;;
 esac
 
+AC_CACHE_CHECK([whether the linker accepts -nopie],
+  [emacs_cv_prog_cc_nopie],
+  [emacs_save_LDFLAGS=$LDFLAGS
+  LDFLAGS="$LDFLAGS -nopie"
+  AC_LINK_IFELSE([AC_LANG_PROGRAM([], [])],
+    [emacs_cv_prog_cc_nopie=yes],
+    [emacs_cv_prog_cc_nopie=no])
+  LDFLAGS=$emacs_save_LDFLAGS])
+if test "$emacs_cv_prog_cc_nopie" = yes; then
+  # Disable PIE to avoid segfaults in temacs (bug#18780)
+  LD_SWITCH_SYSTEM_TEMACS="$LD_SWITCH_SYSTEM_TEMACS -nopie"
+fi
+
 if test x$ac_enable_profiling != x ; then
   case $opsys in
     *freebsd | gnu-linux) ;;
-- 
2.3.2

Message #46 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Stefan Monnier <monnier <at> IRO.UMontreal.CA>
To: Eli Zaretskii <eliz <at> gnu.org>
Cc: ulm <at> gentoo.org, 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Mon, 16 Mar 2015 14:19:43 -0400

> Anyway, please don't make that effective for every supported system;
> e.g., I think it's wrong to do that for Windows, which always produces
> PIE code.

Hmm... so maybe the core of the problem is not PIE and -nopie just ends
up preventing something else (e.g. some kind of address space
randomization of the code).


        Stefan

Message #49 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Stefan Monnier <monnier <at> IRO.UMontreal.CA>
Cc: ulm <at> gentoo.org, 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Mon, 16 Mar 2015 20:29:20 +0200

> From: Stefan Monnier <monnier <at> IRO.UMontreal.CA>
> Cc: ulm <at> gentoo.org, 18784 <at> debbugs.gnu.org
> Date: Mon, 16 Mar 2015 14:19:43 -0400
> 
> > Anyway, please don't make that effective for every supported system;
> > e.g., I think it's wrong to do that for Windows, which always produces
> > PIE code.
> 
> Hmm... so maybe the core of the problem is not PIE and -nopie just ends
> up preventing something else (e.g. some kind of address space
> randomization of the code).

I think we don't really understand why this bug happens.  At least I
have no clear idea of how PIE can cause temacs segfault.

Message #52 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Stefan Monnier <monnier <at> IRO.UMontreal.CA>
To: Eli Zaretskii <eliz <at> gnu.org>
Cc: ulm <at> gentoo.org, 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Mon, 16 Mar 2015 15:06:22 -0400

> I think we don't really understand why this bug happens.  At least I
> have no clear idea of how PIE can cause temacs segfault.

I guess if PIE is used in order to place Emacs's code at randomized
addresses (different address every time an Emacs process is launched),
it can break the dump because our subr objects will end up with pointers
to addresses that aren't valid any more (might be other such problems,
but that's the first that comes to my mind).


        Stefan

Message #55 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Stefan Monnier <monnier <at> IRO.UMontreal.CA>
Cc: ulm <at> gentoo.org, 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Mon, 16 Mar 2015 21:34:34 +0200

> From: Stefan Monnier <monnier <at> IRO.UMontreal.CA>
> Cc: ulm <at> gentoo.org, 18784 <at> debbugs.gnu.org
> Date: Mon, 16 Mar 2015 15:06:22 -0400
> 
> > I think we don't really understand why this bug happens.  At least I
> > have no clear idea of how PIE can cause temacs segfault.
> 
> I guess if PIE is used in order to place Emacs's code at randomized
> addresses (different address every time an Emacs process is launched),
> it can break the dump because our subr objects will end up with pointers
> to addresses that aren't valid any more (might be other such problems,
> but that's the first that comes to my mind).

Then this is only an issue on systems where PIE == address
randomization, isn't it?

Message #58 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Ulrich Mueller <ulm <at> gentoo.org>
To: Eli Zaretskii <eliz <at> gnu.org>
Cc: Stefan Monnier <monnier <at> IRO.UMontreal.CA>, 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Mon, 16 Mar 2015 21:04:08 +0100

>>>>> On Mon, 16 Mar 2015, Eli Zaretskii wrote:

>> I guess if PIE is used in order to place Emacs's code at randomized
>> addresses (different address every time an Emacs process is launched),
>> it can break the dump because our subr objects will end up with pointers
>> to addresses that aren't valid any more (might be other such problems,
>> but that's the first that comes to my mind).

> Then this is only an issue on systems where PIE == address
> randomization, isn't it?

That would be my guess too. Here are some gdb backtraces showing that
the problem occurs in unexelf.c:

   https://bugs.gentoo.org/494316#c13
   https://bugs.gentoo.org/526948#c9
   http://debbugs.gnu.org/13847#5

There is a long history of Emacs catching up with kernel hardening
in this area. It started with setting ADDR_NO_RANDOMIZE via Linux
personality(2), then setting the NORANDEXEC flag with paxctl(1) or
setfattr(1). Now it seems that we need -nopie in addition.
My impression is that these are all workarounds that don't address the
real issue.

Message #63 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Stefan Monnier <monnier <at> IRO.UMontreal.CA>
To: Ulrich Mueller <ulm <at> gentoo.org>
Cc: Eli Zaretskii <eliz <at> gnu.org>, 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Mon, 16 Mar 2015 16:29:48 -0400

> There is a long history of Emacs catching up with kernel hardening
> in this area. It started with setting ADDR_NO_RANDOMIZE via Linux
> personality(2), then setting the NORANDEXEC flag with paxctl(1) or
> setfattr(1). Now it seems that we need -nopie in addition.

But what does "-nopie" mean?  IIUC it means "do not generate PIE code",
so it is a "double-level workaround": not only it doesn't directly fix the
problem we have with randomization but it doesn't directly disable
randomization either.

If OTOH "-nopie" means "indicate that the code should not be relocated
even if it looks like it's position independent", then it's only
a "single-level workaround", like the ADDR_NO_RANDOMIZE and friends.

> My impression is that these are all workarounds that don't address the
> real issue.

AFAIK the only way to address directly the underlying issue is to use
a portable dumper.  Until then we'll have to consider address
randomization as plain bugs that we need to fix with things like
ADDR_NO_RANDOMIZE.


        Stefan "who doesn't really believe in such hardening"

Message #68 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Ulrich Mueller <ulm <at> gentoo.org>
To: 18784 <at> debbugs.gnu.org
Cc: Eli Zaretskii <eliz <at> gnu.org>, Stefan Monnier <monnier <at> iro.umontreal.ca>
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Sun, 29 Mar 2015 14:19:49 +0200

So, how to proceed here? Looks like duplicate bugs are piling up.
Shall I push the patch that I've posted earlier:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=18784#43

Alternatively, the check for the linker accepting -nopie could be
moved to the gnu*) pattern of the case statement above. (The openbsd
case would be left alone then.)

Message #71 received at 18784 <at> debbugs.gnu.org (full text, mbox):

From: Ulrich Mueller <ulm <at> gentoo.org>
To: 18784 <at> debbugs.gnu.org
Subject: Re: bug#18784: Coultdn't compile emacs-24.4
Date: Fri, 3 Apr 2015 13:45:58 +0200

tags 18784 fixed
close 18784 25.1
thanks

> Shall I push the patch that I've posted earlier:
> http://debbugs.gnu.org/cgi/bugreport.cgi?bug=18784#43

Pushed to master.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.