* Build failure: 'seccomp-filter'
@ 2021-04-11 6:45 Colin Baxter
2021-04-11 7:41 ` Eli Zaretskii
0 siblings, 1 reply; 15+ messages in thread
From: Colin Baxter @ 2021-04-11 6:45 UTC (permalink / raw)
To: emacs-devel
Hello,
I get a build with the latest pull of emacs:
Makefile:415: recipe for target 'seccomp-filter' failed
make[1]: *** [seccomp-filter] Error 1
make[1]: Leaving directory '/home/redknight/git/emacs/lib-src'
Makefile:396: recipe for target 'lib-src' failed
make: *** [lib-src] Error 2
Best wishes,
Colin Baxter.
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: Build failure: 'seccomp-filter'
2021-04-11 6:45 Build failure: 'seccomp-filter' Colin Baxter
@ 2021-04-11 7:41 ` Eli Zaretskii
2021-04-11 7:50 ` Colin Baxter
0 siblings, 1 reply; 15+ messages in thread
From: Eli Zaretskii @ 2021-04-11 7:41 UTC (permalink / raw)
To: Colin Baxter; +Cc: emacs-devel
> From: Colin Baxter <m43cap@yandex.com>
> Cc:
> Date: Sun, 11 Apr 2021 07:45:19 +0100
>
> I get a build with the latest pull of emacs:
>
> Makefile:415: recipe for target 'seccomp-filter' failed
> make[1]: *** [seccomp-filter] Error 1
> make[1]: Leaving directory '/home/redknight/git/emacs/lib-src'
> Makefile:396: recipe for target 'lib-src' failed
> make: *** [lib-src] Error 2
Are there no error or warning messages before the "Makefile:415" one?
If there are, please show all the messages.
Thanks.
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: Build failure: 'seccomp-filter'
2021-04-11 7:41 ` Eli Zaretskii
@ 2021-04-11 7:50 ` Colin Baxter
2021-04-11 10:12 ` Philipp Stephani
0 siblings, 1 reply; 15+ messages in thread
From: Colin Baxter @ 2021-04-11 7:50 UTC (permalink / raw)
To: Eli Zaretskii; +Cc: emacs-devel
>>>>> Eli Zaretskii <eliz@gnu.org> writes:
>> From: Colin Baxter <m43cap@yandex.com> Cc: Date: Sun, 11 Apr 2021
>> 07:45:19 +0100
>>
>> I get a build with the latest pull of emacs:
>>
>> Makefile:415: recipe for target 'seccomp-filter' failed make[1]:
>> *** [seccomp-filter] Error 1 make[1]: Leaving directory
>> '/home/redknight/git/emacs/lib-src' Makefile:396: recipe for
>> target 'lib-src' failed make: *** [lib-src] Error 2
> Are there no error or warning messages before the "Makefile:415"
> one? If there are, please show all the messages.
Yes, sorry, I missed the earlier errors. Here is the full list beginning
after the attempt to compile seccomp-filter:
---------- Begin error list ------------
CCLD seccomp-filter
seccomp-filter.c: In function ‘main’:
seccomp-filter.c:142:23: error: ‘SCMP_ACT_KILL_PROCESS’ undeclared (first use in this function)
ctx = seccomp_init (SCMP_ACT_KILL_PROCESS);
^~~~~~~~~~~~~~~~~~~~~
seccomp-filter.c:142:23: note: each undeclared identifier is reported only once for each function it appears in
seccomp-filter.c:151:18: error: ‘SCMP_FLTATR_CTL_LOG’ undeclared (first use in this function)
set_attribute (SCMP_FLTATR_CTL_LOG, 0);
^~~~~~~~~~~~~~~~~~~
In file included from seccomp-filter.c:55:0:
../lib/verify.h:213:41: error: static assertion failed: "verify (sizeof (void *) == 8)"
# define _GL_VERIFY(R, DIAGNOSTIC, ...) _Static_assert (R, DIAGNOSTIC)
^
../lib/verify.h:273:20: note: in expansion of macro ‘_GL_VERIFY’
# define verify(R) _GL_VERIFY (R, "verify (" #R ")", -)
^~~~~~~~~~
seccomp-filter.c:156:3: note: in expansion of macro ‘verify’
verify (sizeof (void *) == 8);
^~~~~~
seccomp-filter.c:169:9: warning: implicit declaration of function ‘SCMP_A2_32’ [-Wimplicit-function-declaration]
SCMP_A2_32 (SCMP_CMP_MASKED_EQ,
^
seccomp-filter.c:105:48: note: in definition of macro ‘RULE’
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^~~~~~~~~~~
seccomp-filter.c:105:20: warning: nested extern declaration of ‘SCMP_A2_32’ [-Wnested-externs]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:168:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (mmap),
^~~~
seccomp-filter.c:174:9: warning: implicit declaration of function ‘SCMP_A3_32’ [-Wimplicit-function-declaration]
SCMP_A3_32 (SCMP_CMP_MASKED_EQ,
^
seccomp-filter.c:105:48: note: in definition of macro ‘RULE’
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^~~~~~~~~~~
seccomp-filter.c:105:20: warning: nested extern declaration of ‘SCMP_A3_32’ [-Wnested-externs]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:168:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (mmap),
^~~~
seccomp-filter.c:105:47: warning: missing braces around initializer [-Wmissing-braces]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:168:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (mmap),
^~~~
seccomp-filter.c:105:47: note: (near initialization for ‘arg_array’)
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:168:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (mmap),
^~~~
seccomp-filter.c:105:47: warning: missing braces around initializer [-Wmissing-braces]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:179:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (mmap),
^~~~
seccomp-filter.c:105:47: note: (near initialization for ‘arg_array’)
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:179:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (mmap),
^~~~
seccomp-filter.c:105:47: warning: missing braces around initializer [-Wmissing-braces]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:189:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (mprotect),
^~~~
seccomp-filter.c:105:47: note: (near initialization for ‘arg_array’)
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:189:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (mprotect),
^~~~
seccomp-filter.c:196:9: warning: implicit declaration of function ‘SCMP_A1_32’ [-Wimplicit-function-declaration]
SCMP_A1_32 (SCMP_CMP_EQ, FUTEX_WAKE_PRIVATE));
^
seccomp-filter.c:105:48: note: in definition of macro ‘RULE’
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^~~~~~~~~~~
seccomp-filter.c:105:20: warning: nested extern declaration of ‘SCMP_A1_32’ [-Wnested-externs]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:195:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (futex),
^~~~
seccomp-filter.c:105:47: warning: missing braces around initializer [-Wmissing-braces]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:195:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (futex),
^~~~
seccomp-filter.c:105:47: note: (near initialization for ‘arg_array’)
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:195:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (futex),
^~~~
seccomp-filter.c:105:47: warning: missing braces around initializer [-Wmissing-braces]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:238:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (open),
^~~~
seccomp-filter.c:105:47: note: (near initialization for ‘arg_array’)
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:238:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (open),
^~~~
seccomp-filter.c:105:47: warning: missing braces around initializer [-Wmissing-braces]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:243:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (openat),
^~~~
seccomp-filter.c:105:47: note: (near initialization for ‘arg_array’)
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:243:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (openat),
^~~~
seccomp-filter.c:251:9: warning: implicit declaration of function ‘SCMP_A0_32’ [-Wimplicit-function-declaration]
SCMP_A0_32 (SCMP_CMP_EQ, STDIN_FILENO),
^
seccomp-filter.c:105:48: note: in definition of macro ‘RULE’
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^~~~~~~~~~~
seccomp-filter.c:105:20: warning: nested extern declaration of ‘SCMP_A0_32’ [-Wnested-externs]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:250:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (ioctl),
^~~~
seccomp-filter.c:105:47: warning: missing braces around initializer [-Wmissing-braces]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:250:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (ioctl),
^~~~
seccomp-filter.c:105:47: note: (near initialization for ‘arg_array’)
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:250:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (ioctl),
^~~~
seccomp-filter.c:105:47: warning: missing braces around initializer [-Wmissing-braces]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:255:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (fcntl),
^~~~
seccomp-filter.c:105:47: note: (near initialization for ‘arg_array’)
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:255:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (fcntl),
^~~~
seccomp-filter.c:105:47: warning: missing braces around initializer [-Wmissing-braces]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:257:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (fcntl64),
^~~~
seccomp-filter.c:105:47: note: (near initialization for ‘arg_array’)
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:257:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (fcntl64),
^~~~
seccomp-filter.c:274:9: warning: implicit declaration of function ‘SCMP_A2_64’ [-Wimplicit-function-declaration]
SCMP_A2_64 (SCMP_CMP_EQ, 0) /* new_limit == NULL */);
^
seccomp-filter.c:105:48: note: in definition of macro ‘RULE’
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^~~~~~~~~~~
seccomp-filter.c:105:20: warning: nested extern declaration of ‘SCMP_A2_64’ [-Wnested-externs]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:272:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (prlimit64),
^~~~
seccomp-filter.c:105:47: warning: missing braces around initializer [-Wmissing-braces]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:272:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (prlimit64),
^~~~
seccomp-filter.c:105:47: note: (near initialization for ‘arg_array’)
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:272:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (prlimit64),
^~~~
seccomp-filter.c:105:47: warning: missing braces around initializer [-Wmissing-braces]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:277:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ERRNO (EPERM), SCMP_SYS (prlimit64),
^~~~
seccomp-filter.c:105:47: note: (near initialization for ‘arg_array’)
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:277:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ERRNO (EPERM), SCMP_SYS (prlimit64),
^~~~
seccomp-filter.c:294:9: warning: implicit declaration of function ‘SCMP_A0_64’ [-Wimplicit-function-declaration]
SCMP_A0_64 (SCMP_CMP_MASKED_EQ,
^
seccomp-filter.c:105:48: note: in definition of macro ‘RULE’
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^~~~~~~~~~~
seccomp-filter.c:105:20: warning: nested extern declaration of ‘SCMP_A0_64’ [-Wnested-externs]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:293:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (clone),
^~~~
seccomp-filter.c:105:47: warning: missing braces around initializer [-Wmissing-braces]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:293:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (clone),
^~~~
seccomp-filter.c:105:47: note: (near initialization for ‘arg_array’)
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:293:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (clone),
^~~~
seccomp-filter.c:105:47: warning: missing braces around initializer [-Wmissing-braces]
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:306:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (prctl),
^~~~
seccomp-filter.c:105:47: note: (near initialization for ‘arg_array’)
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^
seccomp-filter.c:306:3: note: in expansion of macro ‘RULE’
RULE (SCMP_ACT_ALLOW, SCMP_SYS (prctl),
^~~~
Makefile:415: recipe for target 'seccomp-filter' failed
make[1]: *** [seccomp-filter] Error 1
make[1]: Leaving directory '/home/redknight/git/emacs/lib-src'
Makefile:396: recipe for target 'lib-src' failed
make: *** [lib-src] Error 2
--------------- End error list ---------
Best wishes,
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: Build failure: 'seccomp-filter'
2021-04-11 7:50 ` Colin Baxter
@ 2021-04-11 10:12 ` Philipp Stephani
2021-04-11 10:45 ` Colin Baxter
0 siblings, 1 reply; 15+ messages in thread
From: Philipp Stephani @ 2021-04-11 10:12 UTC (permalink / raw)
To: Colin Baxter; +Cc: Eli Zaretskii, Emacs developers
Am So., 11. Apr. 2021 um 09:51 Uhr schrieb Colin Baxter <m43cap@yandex.com>:
>
> >>>>> Eli Zaretskii <eliz@gnu.org> writes:
>
> >> From: Colin Baxter <m43cap@yandex.com> Cc: Date: Sun, 11 Apr 2021
> >> 07:45:19 +0100
> >>
> >> I get a build with the latest pull of emacs:
> >>
> >> Makefile:415: recipe for target 'seccomp-filter' failed make[1]:
> >> *** [seccomp-filter] Error 1 make[1]: Leaving directory
> >> '/home/redknight/git/emacs/lib-src' Makefile:396: recipe for
> >> target 'lib-src' failed make: *** [lib-src] Error 2
>
> > Are there no error or warning messages before the "Makefile:415"
> > one? If there are, please show all the messages.
>
> Yes, sorry, I missed the earlier errors. Here is the full list beginning
> after the attempt to compile seccomp-filter:
>
> ---------- Begin error list ------------
>
> CCLD seccomp-filter
> seccomp-filter.c: In function ‘main’:
> seccomp-filter.c:142:23: error: ‘SCMP_ACT_KILL_PROCESS’ undeclared (first use in this function)
> ctx = seccomp_init (SCMP_ACT_KILL_PROCESS);
> ^~~~~~~~~~~~~~~~~~~~~
This is an interesting failure. Looks like seccomp.h exists on your
system, but doesn't contain the right definitions?
What does your config.log say about seccomp.h (not linux/seccomp.h)
and seccomp_init?
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: Build failure: 'seccomp-filter'
2021-04-11 10:12 ` Philipp Stephani
@ 2021-04-11 10:45 ` Colin Baxter
2021-04-11 14:31 ` Philipp
0 siblings, 1 reply; 15+ messages in thread
From: Colin Baxter @ 2021-04-11 10:45 UTC (permalink / raw)
To: Philipp Stephani; +Cc: Eli Zaretskii, Emacs developers
>>>>> Philipp Stephani <p.stephani2@gmail.com> writes:
> Am So., 11. Apr. 2021 um 09:51 Uhr schrieb Colin Baxter
> <m43cap@yandex.com>:
>>
>> >>>>> Eli Zaretskii <eliz@gnu.org> writes:
>>
>> >> From: Colin Baxter <m43cap@yandex.com> Cc: Date: Sun, 11 Apr
>> 2021 >> 07:45:19 +0100
>> >>
>> >> I get a build with the latest pull of emacs:
>> >>
>> >> Makefile:415: recipe for target 'seccomp-filter' failed
>> make[1]: >> *** [seccomp-filter] Error 1 make[1]: Leaving
>> directory >> '/home/redknight/git/emacs/lib-src' Makefile:396:
>> recipe for >> target 'lib-src' failed make: *** [lib-src] Error 2
>>
>> > Are there no error or warning messages before the
>> "Makefile:415" > one? If there are, please show all the
>> messages.
>>
>> Yes, sorry, I missed the earlier errors. Here is the full list
>> beginning after the attempt to compile seccomp-filter:
>>
>> ---------- Begin error list ------------
>>
>> CCLD seccomp-filter seccomp-filter.c: In function ‘main’:
>> seccomp-filter.c:142:23: error: ‘SCMP_ACT_KILL_PROCESS’
>> undeclared (first use in this function) ctx = seccomp_init
>> (SCMP_ACT_KILL_PROCESS); ^~~~~~~~~~~~~~~~~~~~~
> This is an interesting failure. Looks like seccomp.h exists on
> your system, but doesn't contain the right definitions? What does
> your config.log say about seccomp.h (not linux/seccomp.h) and
> seccomp_init?
I have appended below the occurrences in config.log for seccomp.h. (Do
you want me to attach the whole config.log in a second email?) The only
other mentions are multiple lines of the form
5090:| #define HAVE_LINUX_SECCOMP_H 1
ending with
27758:ac_cv_header_linux_seccomp_h=yes
27771:ac_cv_header_seccomp_h=yes
29332:#define HAVE_LINUX_SECCOMP_H 1
Hope this helps.
---------- Begin relevant portion of config.log -----------
configure:17719: checking linux/seccomp.h usability
configure:17719: gcc -c -g3 -O2 -pthread -isystem /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 conftest.c >&5
configure:17719: $? = 0
configure:17719: result: yes
configure:17719: checking linux/seccomp.h presence
configure:17719: gcc -E conftest.c
configure:17719: $? = 0
configure:17719: result: yes
configure:17719: checking for linux/seccomp.h
configure:17719: result: yes
configure:17731: checking seccomp.h usability
configure:17731: gcc -c -g3 -O2 -pthread -isystem /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 conftest.c >&5
configure:17731: $? = 0
configure:17731: result: yes
configure:17731: checking seccomp.h presence
configure:17731: gcc -E conftest.c
configure:17731: $? = 0
configure:17731: result: yes
configure:17731: checking for seccomp.h
configure:17731: result: yes
configure:17733: checking for seccomp_init in -lseccomp
configure:17758: gcc -o conftest -g3 -O2 -pthread -isystem /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 conftest.c -lseccomp -lX11 -lcairo >&5
configure:17758: $? = 0
configure:17767: result: yes
---------- End relevant portion of config.log ----
Best wishes,
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: Build failure: 'seccomp-filter'
2021-04-11 10:45 ` Colin Baxter
@ 2021-04-11 14:31 ` Philipp
2021-04-11 14:40 ` Philipp Stephani
2021-04-11 15:18 ` Colin Baxter
0 siblings, 2 replies; 15+ messages in thread
From: Philipp @ 2021-04-11 14:31 UTC (permalink / raw)
To: Colin Baxter; +Cc: Eli Zaretskii, Emacs developers
> Am 11.04.2021 um 12:45 schrieb Colin Baxter <m43cap@yandex.com>:
>
>>>>>> Philipp Stephani <p.stephani2@gmail.com> writes:
>
>> Am So., 11. Apr. 2021 um 09:51 Uhr schrieb Colin Baxter
>> <m43cap@yandex.com>:
>>>
>>>>>>>> Eli Zaretskii <eliz@gnu.org> writes:
>>>
>>>>> From: Colin Baxter <m43cap@yandex.com> Cc: Date: Sun, 11 Apr
>>> 2021 >> 07:45:19 +0100
>>>>>
>>>>> I get a build with the latest pull of emacs:
>>>>>
>>>>> Makefile:415: recipe for target 'seccomp-filter' failed
>>> make[1]: >> *** [seccomp-filter] Error 1 make[1]: Leaving
>>> directory >> '/home/redknight/git/emacs/lib-src' Makefile:396:
>>> recipe for >> target 'lib-src' failed make: *** [lib-src] Error 2
>>>
>>>> Are there no error or warning messages before the
>>> "Makefile:415" > one? If there are, please show all the
>>> messages.
>>>
>>> Yes, sorry, I missed the earlier errors. Here is the full list
>>> beginning after the attempt to compile seccomp-filter:
>>>
>>> ---------- Begin error list ------------
>>>
>>> CCLD seccomp-filter seccomp-filter.c: In function ‘main’:
>>> seccomp-filter.c:142:23: error: ‘SCMP_ACT_KILL_PROCESS’
>>> undeclared (first use in this function) ctx = seccomp_init
>>> (SCMP_ACT_KILL_PROCESS); ^~~~~~~~~~~~~~~~~~~~~
>
>
>> This is an interesting failure. Looks like seccomp.h exists on
>> your system, but doesn't contain the right definitions? What does
>> your config.log say about seccomp.h (not linux/seccomp.h) and
>> seccomp_init?
>
> I have appended below the occurrences in config.log for seccomp.h. (Do
> you want me to attach the whole config.log in a second email?) The only
> other mentions are multiple lines of the form
>
> 5090:| #define HAVE_LINUX_SECCOMP_H 1
>
> ending with
>
> 27758:ac_cv_header_linux_seccomp_h=yes
> 27771:ac_cv_header_seccomp_h=yes
> 29332:#define HAVE_LINUX_SECCOMP_H 1
>
> Hope this helps.
>
> ---------- Begin relevant portion of config.log -----------
>
> configure:17719: checking linux/seccomp.h usability
> configure:17719: gcc -c -g3 -O2 -pthread -isystem /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 conftest.c >&5
> configure:17719: $? = 0
> configure:17719: result: yes
> configure:17719: checking linux/seccomp.h presence
> configure:17719: gcc -E conftest.c
> configure:17719: $? = 0
> configure:17719: result: yes
> configure:17719: checking for linux/seccomp.h
> configure:17719: result: yes
> configure:17731: checking seccomp.h usability
> configure:17731: gcc -c -g3 -O2 -pthread -isystem /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 conftest.c >&5
> configure:17731: $? = 0
> configure:17731: result: yes
> configure:17731: checking seccomp.h presence
> configure:17731: gcc -E conftest.c
> configure:17731: $? = 0
> configure:17731: result: yes
> configure:17731: checking for seccomp.h
> configure:17731: result: yes
> configure:17733: checking for seccomp_init in -lseccomp
> configure:17758: gcc -o conftest -g3 -O2 -pthread -isystem /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 conftest.c -lseccomp -lX11 -lcairo >&5
> configure:17758: $? = 0
> configure:17767: result: yes
>
> ---------- End relevant portion of config.log ----
>
>
OK, looks like libseccomp is available and should be working. Does /usr/include/seccomp.h contain at least some of the symbols the compiler complains about?
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: Build failure: 'seccomp-filter'
2021-04-11 14:31 ` Philipp
@ 2021-04-11 14:40 ` Philipp Stephani
2021-04-11 14:52 ` Philipp Stephani
2021-04-11 15:18 ` Colin Baxter
1 sibling, 1 reply; 15+ messages in thread
From: Philipp Stephani @ 2021-04-11 14:40 UTC (permalink / raw)
To: Colin Baxter; +Cc: Eli Zaretskii, Emacs developers
Am So., 11. Apr. 2021 um 16:31 Uhr schrieb Philipp <p.stephani2@gmail.com>:
>
>
>
> > Am 11.04.2021 um 12:45 schrieb Colin Baxter <m43cap@yandex.com>:
> >
> >>>>>> Philipp Stephani <p.stephani2@gmail.com> writes:
> >
> >> Am So., 11. Apr. 2021 um 09:51 Uhr schrieb Colin Baxter
> >> <m43cap@yandex.com>:
> >>>
> >>>>>>>> Eli Zaretskii <eliz@gnu.org> writes:
> >>>
> >>>>> From: Colin Baxter <m43cap@yandex.com> Cc: Date: Sun, 11 Apr
> >>> 2021 >> 07:45:19 +0100
> >>>>>
> >>>>> I get a build with the latest pull of emacs:
> >>>>>
> >>>>> Makefile:415: recipe for target 'seccomp-filter' failed
> >>> make[1]: >> *** [seccomp-filter] Error 1 make[1]: Leaving
> >>> directory >> '/home/redknight/git/emacs/lib-src' Makefile:396:
> >>> recipe for >> target 'lib-src' failed make: *** [lib-src] Error 2
> >>>
> >>>> Are there no error or warning messages before the
> >>> "Makefile:415" > one? If there are, please show all the
> >>> messages.
> >>>
> >>> Yes, sorry, I missed the earlier errors. Here is the full list
> >>> beginning after the attempt to compile seccomp-filter:
> >>>
> >>> ---------- Begin error list ------------
> >>>
> >>> CCLD seccomp-filter seccomp-filter.c: In function ‘main’:
> >>> seccomp-filter.c:142:23: error: ‘SCMP_ACT_KILL_PROCESS’
> >>> undeclared (first use in this function) ctx = seccomp_init
> >>> (SCMP_ACT_KILL_PROCESS); ^~~~~~~~~~~~~~~~~~~~~
> >
> >
> >> This is an interesting failure. Looks like seccomp.h exists on
> >> your system, but doesn't contain the right definitions? What does
> >> your config.log say about seccomp.h (not linux/seccomp.h) and
> >> seccomp_init?
> >
> > I have appended below the occurrences in config.log for seccomp.h. (Do
> > you want me to attach the whole config.log in a second email?) The only
> > other mentions are multiple lines of the form
> >
> > 5090:| #define HAVE_LINUX_SECCOMP_H 1
> >
> > ending with
> >
> > 27758:ac_cv_header_linux_seccomp_h=yes
> > 27771:ac_cv_header_seccomp_h=yes
> > 29332:#define HAVE_LINUX_SECCOMP_H 1
> >
> > Hope this helps.
> >
> > ---------- Begin relevant portion of config.log -----------
> >
> > configure:17719: checking linux/seccomp.h usability
> > configure:17719: gcc -c -g3 -O2 -pthread -isystem /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 conftest.c >&5
> > configure:17719: $? = 0
> > configure:17719: result: yes
> > configure:17719: checking linux/seccomp.h presence
> > configure:17719: gcc -E conftest.c
> > configure:17719: $? = 0
> > configure:17719: result: yes
> > configure:17719: checking for linux/seccomp.h
> > configure:17719: result: yes
> > configure:17731: checking seccomp.h usability
> > configure:17731: gcc -c -g3 -O2 -pthread -isystem /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 conftest.c >&5
> > configure:17731: $? = 0
> > configure:17731: result: yes
> > configure:17731: checking seccomp.h presence
> > configure:17731: gcc -E conftest.c
> > configure:17731: $? = 0
> > configure:17731: result: yes
> > configure:17731: checking for seccomp.h
> > configure:17731: result: yes
> > configure:17733: checking for seccomp_init in -lseccomp
> > configure:17758: gcc -o conftest -g3 -O2 -pthread -isystem /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 conftest.c -lseccomp -lX11 -lcairo >&5
> > configure:17758: $? = 0
> > configure:17767: result: yes
> >
> > ---------- End relevant portion of config.log ----
> >
> >
>
>
> OK, looks like libseccomp is available and should be working. Does /usr/include/seccomp.h contain at least some of the symbols the compiler complains about?
>
Looking at https://github.com/seccomp/libseccomp/blob/main/CHANGELOG,
I guess we need at least version 2.4.0 of libseccomp. I'll see that I
can add a few more checks to configure.ac.
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: Build failure: 'seccomp-filter'
2021-04-11 14:40 ` Philipp Stephani
@ 2021-04-11 14:52 ` Philipp Stephani
2021-04-11 15:40 ` Colin Baxter
2021-04-11 15:55 ` Andreas Schwab
0 siblings, 2 replies; 15+ messages in thread
From: Philipp Stephani @ 2021-04-11 14:52 UTC (permalink / raw)
To: Colin Baxter; +Cc: Eli Zaretskii, Emacs developers
Am So., 11. Apr. 2021 um 16:40 Uhr schrieb Philipp Stephani
<p.stephani2@gmail.com>:
>
> Am So., 11. Apr. 2021 um 16:31 Uhr schrieb Philipp <p.stephani2@gmail.com>:
> >
> >
> >
> > > Am 11.04.2021 um 12:45 schrieb Colin Baxter <m43cap@yandex.com>:
> > >
> > >>>>>> Philipp Stephani <p.stephani2@gmail.com> writes:
> > >
> > >> Am So., 11. Apr. 2021 um 09:51 Uhr schrieb Colin Baxter
> > >> <m43cap@yandex.com>:
> > >>>
> > >>>>>>>> Eli Zaretskii <eliz@gnu.org> writes:
> > >>>
> > >>>>> From: Colin Baxter <m43cap@yandex.com> Cc: Date: Sun, 11 Apr
> > >>> 2021 >> 07:45:19 +0100
> > >>>>>
> > >>>>> I get a build with the latest pull of emacs:
> > >>>>>
> > >>>>> Makefile:415: recipe for target 'seccomp-filter' failed
> > >>> make[1]: >> *** [seccomp-filter] Error 1 make[1]: Leaving
> > >>> directory >> '/home/redknight/git/emacs/lib-src' Makefile:396:
> > >>> recipe for >> target 'lib-src' failed make: *** [lib-src] Error 2
> > >>>
> > >>>> Are there no error or warning messages before the
> > >>> "Makefile:415" > one? If there are, please show all the
> > >>> messages.
> > >>>
> > >>> Yes, sorry, I missed the earlier errors. Here is the full list
> > >>> beginning after the attempt to compile seccomp-filter:
> > >>>
> > >>> ---------- Begin error list ------------
> > >>>
> > >>> CCLD seccomp-filter seccomp-filter.c: In function ‘main’:
> > >>> seccomp-filter.c:142:23: error: ‘SCMP_ACT_KILL_PROCESS’
> > >>> undeclared (first use in this function) ctx = seccomp_init
> > >>> (SCMP_ACT_KILL_PROCESS); ^~~~~~~~~~~~~~~~~~~~~
> > >
> > >
> > >> This is an interesting failure. Looks like seccomp.h exists on
> > >> your system, but doesn't contain the right definitions? What does
> > >> your config.log say about seccomp.h (not linux/seccomp.h) and
> > >> seccomp_init?
> > >
> > > I have appended below the occurrences in config.log for seccomp.h. (Do
> > > you want me to attach the whole config.log in a second email?) The only
> > > other mentions are multiple lines of the form
> > >
> > > 5090:| #define HAVE_LINUX_SECCOMP_H 1
> > >
> > > ending with
> > >
> > > 27758:ac_cv_header_linux_seccomp_h=yes
> > > 27771:ac_cv_header_seccomp_h=yes
> > > 29332:#define HAVE_LINUX_SECCOMP_H 1
> > >
> > > Hope this helps.
> > >
> > > ---------- Begin relevant portion of config.log -----------
> > >
> > > configure:17719: checking linux/seccomp.h usability
> > > configure:17719: gcc -c -g3 -O2 -pthread -isystem /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 conftest.c >&5
> > > configure:17719: $? = 0
> > > configure:17719: result: yes
> > > configure:17719: checking linux/seccomp.h presence
> > > configure:17719: gcc -E conftest.c
> > > configure:17719: $? = 0
> > > configure:17719: result: yes
> > > configure:17719: checking for linux/seccomp.h
> > > configure:17719: result: yes
> > > configure:17731: checking seccomp.h usability
> > > configure:17731: gcc -c -g3 -O2 -pthread -isystem /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 conftest.c >&5
> > > configure:17731: $? = 0
> > > configure:17731: result: yes
> > > configure:17731: checking seccomp.h presence
> > > configure:17731: gcc -E conftest.c
> > > configure:17731: $? = 0
> > > configure:17731: result: yes
> > > configure:17731: checking for seccomp.h
> > > configure:17731: result: yes
> > > configure:17733: checking for seccomp_init in -lseccomp
> > > configure:17758: gcc -o conftest -g3 -O2 -pthread -isystem /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 -isystem /usr/include/cairo -isystem /usr/include/glib-2.0 -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem /usr/include/libpng16 conftest.c -lseccomp -lX11 -lcairo >&5
> > > configure:17758: $? = 0
> > > configure:17767: result: yes
> > >
> > > ---------- End relevant portion of config.log ----
> > >
> > >
> >
> >
> > OK, looks like libseccomp is available and should be working. Does /usr/include/seccomp.h contain at least some of the symbols the compiler complains about?
> >
>
> Looking at https://github.com/seccomp/libseccomp/blob/main/CHANGELOG,
> I guess we need at least version 2.4.0 of libseccomp. I'll see that I
> can add a few more checks to configure.ac.
With commit 725fc96b706c57ef8ceca5e7d82b175d9a72e845, I've now
switched to using pkg-config, which seems cleaner anyway.
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: Build failure: 'seccomp-filter'
2021-04-11 14:31 ` Philipp
2021-04-11 14:40 ` Philipp Stephani
@ 2021-04-11 15:18 ` Colin Baxter
1 sibling, 0 replies; 15+ messages in thread
From: Colin Baxter @ 2021-04-11 15:18 UTC (permalink / raw)
To: Philipp; +Cc: Eli Zaretskii, Emacs developers
>>>>> Philipp <p.stephani2@gmail.com> writes:
>> Am 11.04.2021 um 12:45 schrieb Colin Baxter <m43cap@yandex.com>:
>>
>>>>>>> Philipp Stephani <p.stephani2@gmail.com> writes:
>>
>>> Am So., 11. Apr. 2021 um 09:51 Uhr schrieb Colin Baxter
>>> <m43cap@yandex.com>:
>>>>
>>>>>>>>> Eli Zaretskii <eliz@gnu.org> writes:
>>>>
>>>>> From: Colin Baxter <m43cap@yandex.com> Cc: Date: Sun, 11 Apr 2021
>>>> >> 07:45:19 +0100
>>>>>
>>>>> I get a build with the latest pull of emacs:
>>>>>
>>>>> Makefile:415: recipe for target 'seccomp-filter' failed make[1]:
>>>> >> *** [seccomp-filter] Error 1 make[1]: Leaving directory >>
>>>> '/home/redknight/git/emacs/lib-src' Makefile:396: recipe for >>
>>>> target 'lib-src' failed make: *** [lib-src] Error 2
>>>>
>>>>> Are there no error or warning messages before the
>>>> "Makefile:415" > one? If there are, please show all the
>>>> messages.
>>>>
>>>> Yes, sorry, I missed the earlier errors. Here is the full list
>>>> beginning after the attempt to compile seccomp-filter:
>>>>
>>>> ---------- Begin error list ------------
>>>>
>>>> CCLD seccomp-filter seccomp-filter.c: In function ‘main’:
>>>> seccomp-filter.c:142:23: error: ‘SCMP_ACT_KILL_PROCESS’
>>>> undeclared (first use in this function) ctx = seccomp_init
>>>> (SCMP_ACT_KILL_PROCESS); ^~~~~~~~~~~~~~~~~~~~~
>>
>>
>>> This is an interesting failure. Looks like seccomp.h exists on
>>> your system, but doesn't contain the right definitions? What
>>> does your config.log say about seccomp.h (not linux/seccomp.h)
>>> and seccomp_init?
>>
>> I have appended below the occurrences in config.log for
>> seccomp.h. (Do you want me to attach the whole config.log in a
>> second email?) The only other mentions are multiple lines of the
>> form
>>
>> 5090:| #define HAVE_LINUX_SECCOMP_H 1
>>
>> ending with
>>
>> 27758:ac_cv_header_linux_seccomp_h=yes
>> 27771:ac_cv_header_seccomp_h=yes 29332:#define
>> HAVE_LINUX_SECCOMP_H 1
>>
>> Hope this helps.
>>
>> ---------- Begin relevant portion of config.log -----------
>>
>> configure:17719: checking linux/seccomp.h usability
>> configure:17719: gcc -c -g3 -O2 -pthread -isystem
>> /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0
>> -isystem /usr/include/libpng16 -isystem /usr/include/cairo
>> -isystem /usr/include/glib-2.0 -isystem
>> /usr/lib/i386-linux-gnu/glib-2.0/include -isystem
>> /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem
>> /usr/include/libpng16 -isystem /usr/include/cairo -isystem
>> /usr/include/glib-2.0 -isystem
>> /usr/lib/i386-linux-gnu/glib-2.0/include -isystem
>> /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem
>> /usr/include/libpng16 conftest.c >&5 configure:17719: $? = 0
>> configure:17719: result: yes configure:17719: checking
>> linux/seccomp.h presence configure:17719: gcc -E conftest.c
>> configure:17719: $? = 0 configure:17719: result: yes
>> configure:17719: checking for linux/seccomp.h configure:17719:
>> result: yes configure:17731: checking seccomp.h usability
>> configure:17731: gcc -c -g3 -O2 -pthread -isystem
>> /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0
>> -isystem /usr/include/libpng16 -isystem /usr/include/cairo
>> -isystem /usr/include/glib-2.0 -isystem
>> /usr/lib/i386-linux-gnu/glib-2.0/include -isystem
>> /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem
>> /usr/include/libpng16 -isystem /usr/include/cairo -isystem
>> /usr/include/glib-2.0 -isystem
>> /usr/lib/i386-linux-gnu/glib-2.0/include -isystem
>> /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem
>> /usr/include/libpng16 conftest.c >&5 configure:17731: $? = 0
>> configure:17731: result: yes configure:17731: checking seccomp.h
>> presence configure:17731: gcc -E conftest.c configure:17731: $? =
>> 0 configure:17731: result: yes configure:17731: checking for
>> seccomp.h configure:17731: result: yes configure:17733: checking
>> for seccomp_init in -lseccomp configure:17758: gcc -o conftest
>> -g3 -O2 -pthread -isystem /usr/include/librsvg-2.0 -isystem
>> /usr/include/gdk-pixbuf-2.0 -isystem /usr/include/libpng16
>> -isystem /usr/include/cairo -isystem /usr/include/glib-2.0
>> -isystem /usr/lib/i386-linux-gnu/glib-2.0/include -isystem
>> /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem
>> /usr/include/libpng16 -isystem /usr/include/cairo -isystem
>> /usr/include/glib-2.0 -isystem
>> /usr/lib/i386-linux-gnu/glib-2.0/include -isystem
>> /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem
>> /usr/include/libpng16 conftest.c -lseccomp -lX11 -lcairo >&5
>> configure:17758: $? = 0 configure:17767: result: yes
>>
>> ---------- End relevant portion of config.log ----
>>
>>
> OK, looks like libseccomp is available and should be working.
> Does /usr/include/seccomp.h contain at least some of the symbols
> the compiler complains about?
Well my /usr/include/seccomp.h does contain SCMP_ACT_KILL,
SCMP_CMP_MASKED_EQ, SCMP_ACT_ALLOW, but I don't understand the contents
of the file.
Best wishes, Colin.
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: Build failure: 'seccomp-filter'
2021-04-11 14:52 ` Philipp Stephani
@ 2021-04-11 15:40 ` Colin Baxter
2021-04-11 15:55 ` Andreas Schwab
1 sibling, 0 replies; 15+ messages in thread
From: Colin Baxter @ 2021-04-11 15:40 UTC (permalink / raw)
To: Philipp Stephani; +Cc: Eli Zaretskii, Emacs developers
>>>>> Philipp Stephani <p.stephani2@gmail.com> writes:
> Am So., 11. Apr. 2021 um 16:40 Uhr schrieb Philipp Stephani
> <p.stephani2@gmail.com>:
>>
>> Am So., 11. Apr. 2021 um 16:31 Uhr schrieb Philipp
>> <p.stephani2@gmail.com>:
>> >
>> >
>> >
>> > > Am 11.04.2021 um 12:45 schrieb Colin Baxter
>> <m43cap@yandex.com>:
>> > >
>> > >>>>>> Philipp Stephani <p.stephani2@gmail.com> writes:
>> > >
>> > >> Am So., 11. Apr. 2021 um 09:51 Uhr schrieb Colin Baxter > >>
>> <m43cap@yandex.com>:
>> > >>>
>> > >>>>>>>> Eli Zaretskii <eliz@gnu.org> writes:
>> > >>>
>> > >>>>> From: Colin Baxter <m43cap@yandex.com> Cc: Date: Sun, 11
>> Apr > >>> 2021 >> 07:45:19 +0100
>> > >>>>>
>> > >>>>> I get a build with the latest pull of emacs:
>> > >>>>>
>> > >>>>> Makefile:415: recipe for target 'seccomp-filter' failed >
>> >>> make[1]: >> *** [seccomp-filter] Error 1 make[1]: Leaving >
>> >>> directory >> '/home/redknight/git/emacs/lib-src'
>> Makefile:396: > >>> recipe for >> target 'lib-src' failed make:
>> *** [lib-src] Error 2
>> > >>>
>> > >>>> Are there no error or warning messages before the > >>>
>> "Makefile:415" > one? If there are, please show all the > >>>
>> messages.
>> > >>>
>> > >>> Yes, sorry, I missed the earlier errors. Here is the full
>> list > >>> beginning after the attempt to compile seccomp-filter:
>> > >>>
>> > >>> ---------- Begin error list ------------
>> > >>>
>> > >>> CCLD seccomp-filter seccomp-filter.c: In function ‘main’: >
>> >>> seccomp-filter.c:142:23: error: ‘SCMP_ACT_KILL_PROCESS’ > >>>
>> undeclared (first use in this function) ctx = seccomp_init > >>>
>> (SCMP_ACT_KILL_PROCESS); ^~~~~~~~~~~~~~~~~~~~~
>> > >
>> > >
>> > >> This is an interesting failure. Looks like seccomp.h exists
>> on > >> your system, but doesn't contain the right definitions?
>> What does > >> your config.log say about seccomp.h (not
>> linux/seccomp.h) and > >> seccomp_init?
>> > >
>> > > I have appended below the occurrences in config.log for
>> seccomp.h. (Do > > you want me to attach the whole config.log in
>> a second email?) The only > > other mentions are multiple lines
>> of the form
>> > >
>> > > 5090:| #define HAVE_LINUX_SECCOMP_H 1
>> > >
>> > > ending with
>> > >
>> > > 27758:ac_cv_header_linux_seccomp_h=yes > >
>> 27771:ac_cv_header_seccomp_h=yes > > 29332:#define
>> HAVE_LINUX_SECCOMP_H 1
>> > >
>> > > Hope this helps.
>> > >
>> > > ---------- Begin relevant portion of config.log -----------
>> > >
>> > > configure:17719: checking linux/seccomp.h usability > >
>> configure:17719: gcc -c -g3 -O2 -pthread -isystem > >
>> /usr/include/librsvg-2.0 -isystem /usr/include/gdk-pixbuf-2.0 > >
>> -isystem /usr/include/libpng16 -isystem /usr/include/cairo > >
>> -isystem /usr/include/glib-2.0 -isystem > >
>> /usr/lib/i386-linux-gnu/glib-2.0/include -isystem > >
>> /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem >
>> > /usr/include/libpng16 -isystem /usr/include/cairo -isystem > >
>> /usr/include/glib-2.0 -isystem > >
>> /usr/lib/i386-linux-gnu/glib-2.0/include -isystem > >
>> /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem >
>> > /usr/include/libpng16 conftest.c >&5 > > configure:17719: $? =
>> 0 > > configure:17719: result: yes > > configure:17719: checking
>> linux/seccomp.h presence > > configure:17719: gcc -E conftest.c >
>> > configure:17719: $? = 0 > > configure:17719: result: yes > >
>> configure:17719: checking for linux/seccomp.h > >
>> configure:17719: result: yes > > configure:17731: checking
>> seccomp.h usability > > configure:17731: gcc -c -g3 -O2 -pthread
>> -isystem > > /usr/include/librsvg-2.0 -isystem
>> /usr/include/gdk-pixbuf-2.0 > > -isystem /usr/include/libpng16
>> -isystem /usr/include/cairo > > -isystem /usr/include/glib-2.0
>> -isystem > > /usr/lib/i386-linux-gnu/glib-2.0/include -isystem >
>> > /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem
>> > > /usr/include/libpng16 -isystem /usr/include/cairo -isystem >
>> > /usr/include/glib-2.0 -isystem > >
>> /usr/lib/i386-linux-gnu/glib-2.0/include -isystem > >
>> /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem >
>> > /usr/include/libpng16 conftest.c >&5 > > configure:17731: $? =
>> 0 > > configure:17731: result: yes > > configure:17731: checking
>> seccomp.h presence > > configure:17731: gcc -E conftest.c > >
>> configure:17731: $? = 0 > > configure:17731: result: yes > >
>> configure:17731: checking for seccomp.h > > configure:17731:
>> result: yes > > configure:17733: checking for seccomp_init in
>> -lseccomp > > configure:17758: gcc -o conftest -g3 -O2 -pthread
>> -isystem > > /usr/include/librsvg-2.0 -isystem
>> /usr/include/gdk-pixbuf-2.0 > > -isystem /usr/include/libpng16
>> -isystem /usr/include/cairo > > -isystem /usr/include/glib-2.0
>> -isystem > > /usr/lib/i386-linux-gnu/glib-2.0/include -isystem >
>> > /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem
>> > > /usr/include/libpng16 -isystem /usr/include/cairo -isystem >
>> > /usr/include/glib-2.0 -isystem > >
>> /usr/lib/i386-linux-gnu/glib-2.0/include -isystem > >
>> /usr/include/pixman-1 -isystem /usr/include/freetype2 -isystem >
>> > /usr/include/libpng16 conftest.c -lseccomp -lX11 -lcairo >&5 >
>> > configure:17758: $? = 0 > > configure:17767: result: yes
>> > >
>> > > ---------- End relevant portion of config.log ----
>> > >
>> > >
>> >
>> >
>> > OK, looks like libseccomp is available and should be working.
>> > Does /usr/include/seccomp.h contain at least some of the
>> symbols > the compiler complains about?
>> >
>>
>> Looking at
>> https://github.com/seccomp/libseccomp/blob/main/CHANGELOG, I
>> guess we need at least version 2.4.0 of libseccomp. I'll see that
>> I can add a few more checks to configure.ac.
> With commit 725fc96b706c57ef8ceca5e7d82b175d9a72e845, I've now
> switched to using pkg-config, which seems cleaner anyway.
This seems to be successful. With commit 725fc96b70, I can now build
emacs-28.0.50.
Thanks you.
Best wishes,
Colin Baxter.
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: Build failure: 'seccomp-filter'
2021-04-11 14:52 ` Philipp Stephani
2021-04-11 15:40 ` Colin Baxter
@ 2021-04-11 15:55 ` Andreas Schwab
2021-04-11 17:03 ` Colin Baxter
2021-04-11 18:50 ` Philipp Stephani
1 sibling, 2 replies; 15+ messages in thread
From: Andreas Schwab @ 2021-04-11 15:55 UTC (permalink / raw)
To: Philipp Stephani; +Cc: Colin Baxter, Eli Zaretskii, Emacs developers
There is still something wrong here:
<https://build.opensuse.org/package/live_build_log/home:AndreasSchwab:emacs:master/emacs/15.1/x86_64>
./seccomp-filter \
seccomp-filter.bpf seccomp-filter.pfc
seccomp_init
make[2]: *** [Makefile:428: seccomp-filter.bpf] Error 1
make[2]: *** Waiting for unfinished jobs....
Andreas.
--
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 7578 EB47 D4E5 4D69 2510 2552 DF73 E780 A9DA AEC1
"And now for something completely different."
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: Build failure: 'seccomp-filter'
2021-04-11 15:55 ` Andreas Schwab
@ 2021-04-11 17:03 ` Colin Baxter
2021-04-11 18:50 ` Philipp Stephani
1 sibling, 0 replies; 15+ messages in thread
From: Colin Baxter @ 2021-04-11 17:03 UTC (permalink / raw)
To: Andreas Schwab; +Cc: Philipp Stephani, Eli Zaretskii, Emacs developers
>>>>> Andreas Schwab <schwab@linux-m68k.org> writes:
> There is still something wrong here:
> <https://build.opensuse.org/package/live_build_log/home:AndreasSchwab:emacs:master/emacs/15.1/x86_64>
> ./seccomp-filter \ seccomp-filter.bpf seccomp-filter.pfc
> seccomp_init make[2]: *** [Makefile:428: seccomp-filter.bpf] Error
> 1 make[2]: *** Waiting for unfinished jobs....
I don't see this on Debian. Could it os related?
Best wishes,
Colin Baxter.
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: Build failure: 'seccomp-filter'
2021-04-11 15:55 ` Andreas Schwab
2021-04-11 17:03 ` Colin Baxter
@ 2021-04-11 18:50 ` Philipp Stephani
2021-04-11 22:31 ` Andreas Schwab
1 sibling, 1 reply; 15+ messages in thread
From: Philipp Stephani @ 2021-04-11 18:50 UTC (permalink / raw)
To: Andreas Schwab; +Cc: Colin Baxter, Eli Zaretskii, Emacs developers
Am So., 11. Apr. 2021 um 17:55 Uhr schrieb Andreas Schwab
<schwab@linux-m68k.org>:
>
> There is still something wrong here:
>
> <https://build.opensuse.org/package/live_build_log/home:AndreasSchwab:emacs:master/emacs/15.1/x86_64>
>
> ./seccomp-filter \
> seccomp-filter.bpf seccomp-filter.pfc
> seccomp_init
OK, this is a bit subtle; seccomp_init doesn't work if the kernel
doesn't support the default action. I've now attempted to fix this
with commit 9a57897ea1a125782ff332814d3f978c38162cf8.
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: Build failure: 'seccomp-filter'
2021-04-11 18:50 ` Philipp Stephani
@ 2021-04-11 22:31 ` Andreas Schwab
2021-04-12 7:22 ` Philipp Stephani
0 siblings, 1 reply; 15+ messages in thread
From: Andreas Schwab @ 2021-04-11 22:31 UTC (permalink / raw)
To: Philipp Stephani; +Cc: Colin Baxter, Eli Zaretskii, Emacs developers
There are now different failures:
<https://build.opensuse.org/package/live_build_log/home:AndreasSchwab:emacs:master/emacs/15.2/x86_64>
seccomp-filter.c: In function ‘main’:
seccomp-filter.c:356:34: error: ‘SECCOMP_SET_MODE_FILTER’ undeclared (first use in this function); did you mean ‘HAVE_DECL_SECCOMP_SET_MODE_FILTER’?
SCMP_A0_32 (SCMP_CMP_EQ, SECCOMP_SET_MODE_FILTER),
^
seccomp-filter.c:112:48: note: in definition of macro ‘RULE’
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^~~~~~~~~~~
/usr/include/libseccomp/seccomp.h:211:2: note: in expansion of macro ‘SCMP_CMP64’
SCMP_CMP64(x, y, (uint32_t)(z))
^~~~~~~~~~
/usr/include/libseccomp/seccomp.h:209:2: note: in expansion of macro ‘_SCMP_CMP32_1’
func ## nargs
^~~~
/usr/include/libseccomp/seccomp.h:244:28: note: in expansion of macro ‘SCMP_CMP32’
#define SCMP_A0_32(x, ...) SCMP_CMP32(0, x, __VA_ARGS__)
^~~~~~~~~~
seccomp-filter.c:356:9: note: in expansion of macro ‘SCMP_A0_32’
SCMP_A0_32 (SCMP_CMP_EQ, SECCOMP_SET_MODE_FILTER),
^~~~~~~~~~
seccomp-filter.c:356:34: note: each undeclared identifier is reported only once for each function it appears in
SCMP_A0_32 (SCMP_CMP_EQ, SECCOMP_SET_MODE_FILTER),
^
seccomp-filter.c:112:48: note: in definition of macro ‘RULE’
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^~~~~~~~~~~
/usr/include/libseccomp/seccomp.h:211:2: note: in expansion of macro ‘SCMP_CMP64’
SCMP_CMP64(x, y, (uint32_t)(z))
^~~~~~~~~~
/usr/include/libseccomp/seccomp.h:209:2: note: in expansion of macro ‘_SCMP_CMP32_1’
func ## nargs
^~~~
/usr/include/libseccomp/seccomp.h:244:28: note: in expansion of macro ‘SCMP_CMP32’
#define SCMP_A0_32(x, ...) SCMP_CMP32(0, x, __VA_ARGS__)
^~~~~~~~~~
seccomp-filter.c:356:9: note: in expansion of macro ‘SCMP_A0_32’
SCMP_A0_32 (SCMP_CMP_EQ, SECCOMP_SET_MODE_FILTER),
^~~~~~~~~~
seccomp-filter.c:357:34: error: ‘SECCOMP_FILTER_FLAG_TSYNC’ undeclared (first use in this function); did you mean ‘SCMP_FLTATR_CTL_TSYNC’?
SCMP_A1_32 (SCMP_CMP_EQ, SECCOMP_FILTER_FLAG_TSYNC));
^
seccomp-filter.c:112:48: note: in definition of macro ‘RULE’
const struct scmp_arg_cmp arg_array[] = {__VA_ARGS__}; \
^~~~~~~~~~~
/usr/include/libseccomp/seccomp.h:211:2: note: in expansion of macro ‘SCMP_CMP64’
SCMP_CMP64(x, y, (uint32_t)(z))
^~~~~~~~~~
/usr/include/libseccomp/seccomp.h:209:2: note: in expansion of macro ‘_SCMP_CMP32_1’
func ## nargs
^~~~
/usr/include/libseccomp/seccomp.h:255:28: note: in expansion of macro ‘SCMP_CMP32’
#define SCMP_A1_32(x, ...) SCMP_CMP32(1, x, __VA_ARGS__)
^~~~~~~~~~
seccomp-filter.c:357:9: note: in expansion of macro ‘SCMP_A1_32’
SCMP_A1_32 (SCMP_CMP_EQ, SECCOMP_FILTER_FLAG_TSYNC));
^~~~~~~~~~
Andreas.
--
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 7578 EB47 D4E5 4D69 2510 2552 DF73 E780 A9DA AEC1
"And now for something completely different."
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: Build failure: 'seccomp-filter'
2021-04-11 22:31 ` Andreas Schwab
@ 2021-04-12 7:22 ` Philipp Stephani
0 siblings, 0 replies; 15+ messages in thread
From: Philipp Stephani @ 2021-04-12 7:22 UTC (permalink / raw)
To: Andreas Schwab; +Cc: Colin Baxter, Eli Zaretskii, Emacs developers
Am Mo., 12. Apr. 2021 um 00:31 Uhr schrieb Andreas Schwab
<schwab@linux-m68k.org>:
>
> There are now different failures:
>
> <https://build.opensuse.org/package/live_build_log/home:AndreasSchwab:emacs:master/emacs/15.2/x86_64>
>
> seccomp-filter.c: In function ‘main’:
> seccomp-filter.c:356:34: error: ‘SECCOMP_SET_MODE_FILTER’ undeclared (first use in this function); did you mean ‘HAVE_DECL_SECCOMP_SET_MODE_FILTER’?
> SCMP_A0_32 (SCMP_CMP_EQ, SECCOMP_SET_MODE_FILTER),
> ^
I've tried to address this with commit 17d20bb3cb.
^ permalink raw reply [flat|nested] 15+ messages in thread
end of thread, other threads:[~2021-04-12 7:22 UTC | newest]
Thread overview: 15+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2021-04-11 6:45 Build failure: 'seccomp-filter' Colin Baxter
2021-04-11 7:41 ` Eli Zaretskii
2021-04-11 7:50 ` Colin Baxter
2021-04-11 10:12 ` Philipp Stephani
2021-04-11 10:45 ` Colin Baxter
2021-04-11 14:31 ` Philipp
2021-04-11 14:40 ` Philipp Stephani
2021-04-11 14:52 ` Philipp Stephani
2021-04-11 15:40 ` Colin Baxter
2021-04-11 15:55 ` Andreas Schwab
2021-04-11 17:03 ` Colin Baxter
2021-04-11 18:50 ` Philipp Stephani
2021-04-11 22:31 ` Andreas Schwab
2021-04-12 7:22 ` Philipp Stephani
2021-04-11 15:18 ` Colin Baxter
Code repositories for project(s) associated with this public inbox
https://git.savannah.gnu.org/cgit/emacs.git
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).