| Message ID | 20230527133252.284132-2-f_l_k@t-online.de |
|---|---|
| State | Accepted, archived |
| Commit | c6198e1b1c3e3a3413b6ff9f014b40114f1850db |
| Headers | show |
| Series | [1/4] ell: upgrade 0.56 -> 0.57 | expand |
Can you please show how it looks like? We do have a test for host contamination via headers, so I'd like to see what happens exactly and where the test isn't catching it. Same for patches 3/4, 4/4. Alex On Sat, 27 May 2023 at 15:33, Markus Volk <f_l_k@t-online.de> wrote: > > crypt.h is otherwise taken from the host machine > > Signed-off-by: Markus Volk <f_l_k@t-online.de> > --- > meta/recipes-devtools/python/python3_3.11.2.bb | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/meta/recipes-devtools/python/python3_3.11.2.bb b/meta/recipes-devtools/python/python3_3.11.2.bb > index 421a305e22..1ca4ad35be 100644 > --- a/meta/recipes-devtools/python/python3_3.11.2.bb > +++ b/meta/recipes-devtools/python/python3_3.11.2.bb > @@ -72,7 +72,7 @@ ALTERNATIVE_LINK_NAME[python3-config] = "${bindir}/python${PYTHON_MAJMIN}-config > ALTERNATIVE_TARGET[python3-config] = "${bindir}/python${PYTHON_MAJMIN}-config-${MULTILIB_SUFFIX}" > > > -DEPENDS = "bzip2-replacement-native expat libffi bzip2 openssl sqlite3 zlib virtual/libintl xz virtual/crypt util-linux-libuuid libtirpc libnsl2 autoconf-archive-native ncurses" > +DEPENDS = "bzip2-replacement-native expat libffi bzip2 openssl sqlite3 zlib virtual/libintl xz virtual/crypt util-linux-libuuid libtirpc libnsl2 autoconf-archive-native ncurses libxcrypt-native" > DEPENDS:append:class-target = " python3-native" > DEPENDS:append:class-nativesdk = " python3-native" > > -- > 2.34.1 > > > -=-=-=-=-=-=-=-=-=-=-=- > Links: You receive all messages sent to this group. > View/Reply Online (#181806): https://lists.openembedded.org/g/openembedded-core/message/181806 > Mute This Topic: https://lists.openembedded.org/mt/99167320/1686489 > Group Owner: openembedded-core+owner@lists.openembedded.org > Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [alex.kanavin@gmail.com] > -=-=-=-=-=-=-=-=-=-=-=- >
Also, how is the native header used in the target build? Isn't that incorrect in itself? Alex On Sun, 28 May 2023 at 15:40, Alexander Kanavin via lists.openembedded.org <alex.kanavin=gmail.com@lists.openembedded.org> wrote: > > Can you please show how it looks like? We do have a test for host > contamination via headers, so I'd like to see what happens exactly and > where the test isn't catching it. > > Same for patches 3/4, 4/4. > > Alex > > On Sat, 27 May 2023 at 15:33, Markus Volk <f_l_k@t-online.de> wrote: > > > > crypt.h is otherwise taken from the host machine > > > > Signed-off-by: Markus Volk <f_l_k@t-online.de> > > --- > > meta/recipes-devtools/python/python3_3.11.2.bb | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/meta/recipes-devtools/python/python3_3.11.2.bb b/meta/recipes-devtools/python/python3_3.11.2.bb > > index 421a305e22..1ca4ad35be 100644 > > --- a/meta/recipes-devtools/python/python3_3.11.2.bb > > +++ b/meta/recipes-devtools/python/python3_3.11.2.bb > > @@ -72,7 +72,7 @@ ALTERNATIVE_LINK_NAME[python3-config] = "${bindir}/python${PYTHON_MAJMIN}-config > > ALTERNATIVE_TARGET[python3-config] = "${bindir}/python${PYTHON_MAJMIN}-config-${MULTILIB_SUFFIX}" > > > > > > -DEPENDS = "bzip2-replacement-native expat libffi bzip2 openssl sqlite3 zlib virtual/libintl xz virtual/crypt util-linux-libuuid libtirpc libnsl2 autoconf-archive-native ncurses" > > +DEPENDS = "bzip2-replacement-native expat libffi bzip2 openssl sqlite3 zlib virtual/libintl xz virtual/crypt util-linux-libuuid libtirpc libnsl2 autoconf-archive-native ncurses libxcrypt-native" > > DEPENDS:append:class-target = " python3-native" > > DEPENDS:append:class-nativesdk = " python3-native" > > > > -- > > 2.34.1 > > > > > > > > > > -=-=-=-=-=-=-=-=-=-=-=- > Links: You receive all messages sent to this group. > View/Reply Online (#181824): https://lists.openembedded.org/g/openembedded-core/message/181824 > Mute This Topic: https://lists.openembedded.org/mt/99167320/1686489 > Group Owner: openembedded-core+owner@lists.openembedded.org > Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [alex.kanavin@gmail.com] > -=-=-=-=-=-=-=-=-=-=-=- >
I noticed this because I was trying to see if my image could reproduce itself. To do this I installed some build tools, perl and python modules and tried if a core-image-minimal can be built in this environment. This failed for some packages because crypt() was missing. I did not have any dev packages installed, so my guess was that it was because my image was missing crypt.h. To reproduce this on my host machine (Ubuntu 22.04), I ran sudo mv /usr/include/crypt.h /usr/include/crypt.h.old sudo rm -r tmp sstate-cache Now I also had errors on my host system Am So, 28. Mai 2023 um 15:39:47 +0200 schrieb Alexander Kanavin <alex.kanavin@gmail.com>: > Can you please show how it looks like? We do have a test for host > contamination via headers, so I'd like to see what happens exactly and > where the test isn't catching it. > > Same for patches 3/4, 4/4. > > Alex > > On Sat, 27 May 2023 at 15:33, Markus Volk <f_l_k@t-online.de > <mailto:f_l_k@t-online.de>> wrote: >> >> crypt.h is otherwise taken from the host machine >> >> Signed-off-by: Markus Volk <f_l_k@t-online.de >> <mailto:f_l_k@t-online.de>> >> --- >> meta/recipes-devtools/python/python3_3.11.2.bb | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/meta/recipes-devtools/python/python3_3.11.2.bb >> b/meta/recipes-devtools/python/python3_3.11.2.bb >> index 421a305e22..1ca4ad35be 100644 >> --- a/meta/recipes-devtools/python/python3_3.11.2.bb >> +++ b/meta/recipes-devtools/python/python3_3.11.2.bb >> @@ -72,7 +72,7 @@ ALTERNATIVE_LINK_NAME[python3-config] = >> "${bindir}/python${PYTHON_MAJMIN}-config >> ALTERNATIVE_TARGET[python3-config] = >> "${bindir}/python${PYTHON_MAJMIN}-config-${MULTILIB_SUFFIX}" >> >> >> -DEPENDS = "bzip2-replacement-native expat libffi bzip2 openssl >> sqlite3 zlib virtual/libintl xz virtual/crypt util-linux-libuuid >> libtirpc libnsl2 autoconf-archive-native ncurses" >> +DEPENDS = "bzip2-replacement-native expat libffi bzip2 openssl >> sqlite3 zlib virtual/libintl xz virtual/crypt util-linux-libuuid >> libtirpc libnsl2 autoconf-archive-native ncurses libxcrypt-native" >> DEPENDS:append:class-target = " python3-native" >> DEPENDS:append:class-nativesdk = " python3-native" >> >> -- >> 2.34.1 >> >> >> >> > > -=-=-=-=-=-=-=-=-=-=-=- > Links: You receive all messages sent to this group. > View/Reply Online (#181824): > <https://lists.openembedded.org/g/openembedded-core/message/181824> > Mute This Topic: <https://lists.openembedded.org/mt/99167320/3618223> > Group Owner: openembedded-core+owner@lists.openembedded.org > <mailto:openembedded-core+owner@lists.openembedded.org> > Unsubscribe: > <https://lists.openembedded.org/g/openembedded-core/unsub> > [f_l_k@t-online.de <mailto:f_l_k@t-online.de>] > -=-=-=-=-=-=-=-=-=-=-=- >
Am So, 28. Mai 2023 um 15:41:10 +0200 schrieb Alexander Kanavin <alex.kanavin@gmail.com>: > Also, how is the native header used in the target build? Isn't that > incorrect in itself? I guess its only needed to build the native variants of that packages
But can you show the actual error please? If it’s a native only issue then a different fix is needed, so that native dependency isn’t added to a target build. Alex On Sun 28. May 2023 at 16.03, Markus Volk <f_l_k@t-online.de> wrote: > Am So, 28. Mai 2023 um 15:41:10 +0200 schrieb Alexander Kanavin < > alex.kanavin@gmail.com>: > > Also, how is the native header used in the target build? Isn't that > incorrect in itself? > > > I guess its only needed to build the native variants of that packages >
Am So, 28. Mai 2023 um 16:23:36 +0200 schrieb Alexander Kanavin
<alex.kanavin@gmail.com>:
> But can you show the actual error please?
python3-native for example fails like this:
ERROR: python3-native-3.11.2-r0 do_install:
ExecutionError('/home/flk/poky/build/tmp/work/x86_64-linux/python3-native/3.11.2-r0/temp/run.do_install.4082972',
1, None, None)
ERROR: Logfile of failure stored in:
/home/flk/poky/build/tmp/work/x86_64-linux/python3-native/3.11.2-r0/temp/log.do_install.4082972
Log data follows:
| DEBUG: Executing python function autotools_aclocals
| DEBUG: SITE files ['endian-little', 'common-linux', 'common-glibc',
'bit-64', 'x86_64-linux', 'common']
| DEBUG: Python function autotools_aclocals finished
| DEBUG: Executing shell function do_install
| The necessary bits to build these optional modules were not found:
| _crypt
|
| WARNING: exit code 1 from a shell command.
ERROR: Task
(virtual:native:/home/flk/poky/meta/recipes-devtools/python/python3_3.11.2.bb:do_install)
failed with exit code '1'
NOTE: Tasks Summary: Attempted 1089 tasks of which 1078 didn't need to
be rerun and 1 failed.
Summary: 1 task failed:
virtual:native:/home/flk/poky/meta/recipes-devtools/python/python3_3.11.2.bb:do_install
Summary: There was 1 ERROR message, returning a non-zero exit code.
ruby-native fails like this:
compiling ../ruby-3.2.2/vm_trace.c
| In file included from ../ruby-3.2.2/include/ruby/ruby.h:42,
| from ../ruby-3.2.2/debug_counter.h:362,
| from ../ruby-3.2.2/string.c:24:
| ../ruby-3.2.2/string.c: In function ‘rb_str_crypt’:
| ../ruby-3.2.2/string.c:10169:35: error: invalid application of
‘sizeof’ to incomplete type ‘struct crypt_data’
| 10169 | data = ALLOCV(databuf, sizeof(struct crypt_data));
| | ^~~~~~
| ../ruby-3.2.2/include/ruby/internal/memory.h:299:7: note: in
definition of macro ‘RB_ALLOCV’
| 299 | ((n) < RUBY_ALLOCV_LIMIT ? \
| | ^
| ../ruby-3.2.2/string.c:10169:12: note: in expansion of macro
‘ALLOCV’
| 10169 | data = ALLOCV(databuf, sizeof(struct crypt_data));
| | ^~~~~~
| compiling ../ruby-3.2.2/missing/setproctitle.c
| compiling ../ruby-3.2.2/missing/strlcpy.c
| compiling ../ruby-3.2.2/missing/strlcat.c
| In file included from /usr/include/stdlib.h:574,
| from ../ruby-3.2.2/include/ruby/defines.h:27,
| from ../ruby-3.2.2/include/ruby/ruby.h:25,
| from ../ruby-3.2.2/debug_counter.h:362,
| from ../ruby-3.2.2/string.c:24:
| ../ruby-3.2.2/string.c:10169:35: error: invalid application of
‘sizeof’ to incomplete type ‘struct crypt_data’
| 10169 | data = ALLOCV(databuf, sizeof(struct crypt_data));
| | ^~~~~~
| ../ruby-3.2.2/include/ruby/internal/memory.h:398:20: note: in
expansion of macro ‘RB_ALLOCV’
| 398 | #define ALLOCV RB_ALLOCV /**< @old{RB_ALLOCV} */
| | ^~~~~~~~~
| ../ruby-3.2.2/string.c:10169:12: note: in expansion of macro
‘ALLOCV’
| 10169 | data = ALLOCV(databuf, sizeof(struct crypt_data));
| | ^~~~~~
| In file included from ../ruby-3.2.2/include/ruby/ruby.h:42,
| from ../ruby-3.2.2/debug_counter.h:362,
| from ../ruby-3.2.2/string.c:24:
| ../ruby-3.2.2/string.c:10169:35: error: invalid application of
‘sizeof’ to incomplete type ‘struct crypt_data’
| 10169 | data = ALLOCV(databuf, sizeof(struct crypt_data));
| | ^~~~~~
| ../ruby-3.2.2/include/ruby/internal/memory.h:301:33: note: in
definition of macro ‘RB_ALLOCV’
| 301 | rb_alloc_tmp_buffer(&(v), (n)))
| | ^
| ../ruby-3.2.2/string.c:10169:12: note: in expansion of macro
‘ALLOCV’
| 10169 | data = ALLOCV(databuf, sizeof(struct crypt_data));
| | ^~~~~~
| ../ruby-3.2.2/string.c:10173:11: warning: implicit declaration of
function ‘crypt_r’; did you mean ‘crypt’?
[-Wimplicit-function-declaration]
| 10173 | res = crypt_r(s, saltp, data);
| | ^~~~~~~
| | crypt
| ../ruby-3.2.2/string.c:10173:9: warning: assignment to ‘char *’
from ‘int’ makes pointer from integer without a cast
[-Wint-conversion]
| 10173 | res = crypt_r(s, saltp, data);
| | ^
Thanks, then all three recipes should add libxcrypt to DEPENDS (not libxcrypt-native). For native variants it will be adjusted automatically. Alex On Sun, 28 May 2023 at 16:53, Markus Volk <f_l_k@t-online.de> wrote: > > Am So, 28. Mai 2023 um 16:23:36 +0200 schrieb Alexander Kanavin <alex.kanavin@gmail.com>: > > But can you show the actual error please? > > > python3-native for example fails like this: > ERROR: python3-native-3.11.2-r0 do_install: ExecutionError('/home/flk/poky/build/tmp/work/x86_64-linux/python3-native/3.11.2-r0/temp/run.do_install.4082972', 1, None, None) > ERROR: Logfile of failure stored in: /home/flk/poky/build/tmp/work/x86_64-linux/python3-native/3.11.2-r0/temp/log.do_install.4082972 > Log data follows: > | DEBUG: Executing python function autotools_aclocals > | DEBUG: SITE files ['endian-little', 'common-linux', 'common-glibc', 'bit-64', 'x86_64-linux', 'common'] > | DEBUG: Python function autotools_aclocals finished > | DEBUG: Executing shell function do_install > | The necessary bits to build these optional modules were not found: > | _crypt > | > | WARNING: exit code 1 from a shell command. > ERROR: Task (virtual:native:/home/flk/poky/meta/recipes-devtools/python/python3_3.11.2.bb:do_install) failed with exit code '1' > NOTE: Tasks Summary: Attempted 1089 tasks of which 1078 didn't need to be rerun and 1 failed. > > Summary: 1 task failed: > virtual:native:/home/flk/poky/meta/recipes-devtools/python/python3_3.11.2.bb:do_install > Summary: There was 1 ERROR message, returning a non-zero exit code. > > > ruby-native fails like this: > > compiling ../ruby-3.2.2/vm_trace.c > | In file included from ../ruby-3.2.2/include/ruby/ruby.h:42, > | from ../ruby-3.2.2/debug_counter.h:362, > | from ../ruby-3.2.2/string.c:24: > | ../ruby-3.2.2/string.c: In function ‘rb_str_crypt’: > | ../ruby-3.2.2/string.c:10169:35: error: invalid application of ‘sizeof’ to incomplete type ‘struct crypt_data’ > | 10169 | data = ALLOCV(databuf, sizeof(struct crypt_data)); > | | ^~~~~~ > | ../ruby-3.2.2/include/ruby/internal/memory.h:299:7: note: in definition of macro ‘RB_ALLOCV’ > | 299 | ((n) < RUBY_ALLOCV_LIMIT ? \ > | | ^ > | ../ruby-3.2.2/string.c:10169:12: note: in expansion of macro ‘ALLOCV’ > | 10169 | data = ALLOCV(databuf, sizeof(struct crypt_data)); > | | ^~~~~~ > | compiling ../ruby-3.2.2/missing/setproctitle.c > | compiling ../ruby-3.2.2/missing/strlcpy.c > | compiling ../ruby-3.2.2/missing/strlcat.c > | In file included from /usr/include/stdlib.h:574, > | from ../ruby-3.2.2/include/ruby/defines.h:27, > | from ../ruby-3.2.2/include/ruby/ruby.h:25, > | from ../ruby-3.2.2/debug_counter.h:362, > | from ../ruby-3.2.2/string.c:24: > | ../ruby-3.2.2/string.c:10169:35: error: invalid application of ‘sizeof’ to incomplete type ‘struct crypt_data’ > | 10169 | data = ALLOCV(databuf, sizeof(struct crypt_data)); > | | ^~~~~~ > | ../ruby-3.2.2/include/ruby/internal/memory.h:398:20: note: in expansion of macro ‘RB_ALLOCV’ > | 398 | #define ALLOCV RB_ALLOCV /**< @old{RB_ALLOCV} */ > | | ^~~~~~~~~ > | ../ruby-3.2.2/string.c:10169:12: note: in expansion of macro ‘ALLOCV’ > | 10169 | data = ALLOCV(databuf, sizeof(struct crypt_data)); > | | ^~~~~~ > | In file included from ../ruby-3.2.2/include/ruby/ruby.h:42, > | from ../ruby-3.2.2/debug_counter.h:362, > | from ../ruby-3.2.2/string.c:24: > | ../ruby-3.2.2/string.c:10169:35: error: invalid application of ‘sizeof’ to incomplete type ‘struct crypt_data’ > | 10169 | data = ALLOCV(databuf, sizeof(struct crypt_data)); > | | ^~~~~~ > | ../ruby-3.2.2/include/ruby/internal/memory.h:301:33: note: in definition of macro ‘RB_ALLOCV’ > | 301 | rb_alloc_tmp_buffer(&(v), (n))) > | | ^ > | ../ruby-3.2.2/string.c:10169:12: note: in expansion of macro ‘ALLOCV’ > | 10169 | data = ALLOCV(databuf, sizeof(struct crypt_data)); > | | ^~~~~~ > | ../ruby-3.2.2/string.c:10173:11: warning: implicit declaration of function ‘crypt_r’; did you mean ‘crypt’? [-Wimplicit-function-declaration] > | 10173 | res = crypt_r(s, saltp, data); > | | ^~~~~~~ > | | crypt > | ../ruby-3.2.2/string.c:10173:9: warning: assignment to ‘char *’ from ‘int’ makes pointer from integer without a cast [-Wint-conversion] > | 10173 | res = crypt_r(s, saltp, data); > | | ^ > >
Am So, 28. Mai 2023 um 17:01:44 +0200 schrieb Alexander Kanavin
<alex.kanavin@gmail.com>:
> Thanks, then all three recipes should add libxcrypt to DEPENDS
But wouldn't that lead to multiple virtual/crypt providers on musl ?
Adding libxcrypt-native to DEPENDS definitely isn't correct, so you could check why virtual/crypt dependency isn't working properly for adding the needed headers into native variant's sysroot. If all fails, adjust DEPENDS:append:class-native. Alex On Sun, 28 May 2023 at 17:08, Markus Volk <f_l_k@t-online.de> wrote: > > Am So, 28. Mai 2023 um 17:01:44 +0200 schrieb Alexander Kanavin <alex.kanavin@gmail.com>: > > Thanks, then all three recipes should add libxcrypt to DEPENDS > > > But wouldn't that lead to multiple virtual/crypt providers on musl ?
On Sun, 2023-05-28 at 17:16 +0200, Alexander Kanavin wrote: > Adding libxcrypt-native to DEPENDS definitely isn't correct, so you > could check why virtual/crypt dependency isn't working properly for > adding the needed headers into native variant's sysroot. If all fails, > adjust DEPENDS:append:class-native. Alex is right and I should have spotted this, the -native dependency definitely isn't right. We'll either need to fix things or revert them and wait for the right fix... Cheers, Richard
On Mon, 2023-05-29 at 14:02 +0200, Markus Volk wrote: > Am Mo, 29. Mai 2023 um 08:08:09 +0100 schrieb Richard Purdie > <richard.purdie@linuxfoundation.org>: > > Alex is right and I should have spotted this, the -native > > dependency > > definitely isn't right. We'll either need to fix things or revert > > them > > and wait for the right fix... > > > Absolutely. I had a quick look at native.bbclass yesterday, but > haven't found a nice way to deal with this yet. This week I'm on > night shift and therefore have rather little time to deal with the > problem. I can send v2 patches which just add > > DEPENDS:append:class-native = " libxcrypt-native" > > But revert would also be fine for me, as this is more of a niche > problem I'm going to revert those changes and then we can work out the correct fix without any time pressures (and then I won't need to keep this on my 'need to do something about' mental checklist!). Cheers, Richard
Am Mo, 29. Mai 2023 um 13:34:40 +0100 schrieb Richard Purdie <richard.purdie@linuxfoundation.org>: > I'm going to revert those changes and then we can work out the correct > fix without any time pressures (and then I won't need to keep this on > my 'need to do something about' mental checklist!). I had some time to look at this problem. virtual/crypt-native is not installed into recipe-sysroot-native because of this entry in bitbake.conf: ASSUME_PROVIDED = "virtual/crypt-native" Removing the entry solved the issue for me. I could successfully build core-image-minimal for musl and glibc without crypt.h from host machine. But I don't know the reason why this is ASSUME_PROVIDED, so I'm unsure if this should even be touched. In a second step I created a musl-native recipe to be able to install musl crypt.h into recipe-sysroot-native, but I'm not sure if this is necessary either. <https://github.com/MarkusVolk/openembedded-core/commits/virtual/crypt-native>
On Sat, 3 Jun 2023 at 19:52, Markus Volk <f_l_k@t-online.de> wrote: > I had some time to look at this problem. virtual/crypt-native is not installed into recipe-sysroot-native because of this entry in bitbake.conf: > ASSUME_PROVIDED = "virtual/crypt-native" > > Removing the entry solved the issue for me. I could successfully build core-image-minimal for musl and glibc without crypt.h from host machine. > > But I don't know the reason why this is ASSUME_PROVIDED, so I'm unsure if this should even be touched. > > In a second step I created a musl-native recipe to be able to install musl crypt.h into recipe-sysroot-native, but I'm not sure if this is necessary either. > > https://github.com/MarkusVolk/openembedded-core/commits/virtual/crypt-native I took a look; I believe this is by design, and for native recipes crypt.h does indeed need to be provided by the host distro, and not any of the recipes. This is where it was added: https://git.yoctoproject.org/poky/commit/?id=29f65bda6d2c Alex
Thanks for the link Am Sa, 3. Jun 2023 um 20:20:48 +0200 schrieb Alexander Kanavin <alex.kanavin@gmail.com>: > On Sat, 3 Jun 2023 at 19:52, Markus Volk <f_l_k@t-online.de > <mailto:f_l_k@t-online.de>> wrote: >> I had some time to look at this problem. virtual/crypt-native is >> not installed into recipe-sysroot-native because of this entry in >> bitbake.conf: >> ASSUME_PROVIDED = "virtual/crypt-native" >> >> Removing the entry solved the issue for me. I could successfully >> build core-image-minimal for musl and glibc without crypt.h from >> host machine. >> >> But I don't know the reason why this is ASSUME_PROVIDED, so I'm >> unsure if this should even be touched. >> >> In a second step I created a musl-native recipe to be able to >> install musl crypt.h into recipe-sysroot-native, but I'm not sure if >> this is necessary either. >> >> >> <https://github.com/MarkusVolk/openembedded-core/commits/virtual/crypt-native> > > I took a look; I believe this is by design, and for native recipes > crypt.h does indeed need to be provided by the host distro, and not > any of the recipes. This is where it was added: > <https://git.yoctoproject.org/poky/commit/?id=29f65bda6d2c> > > Alex
diff --git a/meta/recipes-devtools/python/python3_3.11.2.bb b/meta/recipes-devtools/python/python3_3.11.2.bb index 421a305e22..1ca4ad35be 100644 --- a/meta/recipes-devtools/python/python3_3.11.2.bb +++ b/meta/recipes-devtools/python/python3_3.11.2.bb @@ -72,7 +72,7 @@ ALTERNATIVE_LINK_NAME[python3-config] = "${bindir}/python${PYTHON_MAJMIN}-config ALTERNATIVE_TARGET[python3-config] = "${bindir}/python${PYTHON_MAJMIN}-config-${MULTILIB_SUFFIX}" -DEPENDS = "bzip2-replacement-native expat libffi bzip2 openssl sqlite3 zlib virtual/libintl xz virtual/crypt util-linux-libuuid libtirpc libnsl2 autoconf-archive-native ncurses" +DEPENDS = "bzip2-replacement-native expat libffi bzip2 openssl sqlite3 zlib virtual/libintl xz virtual/crypt util-linux-libuuid libtirpc libnsl2 autoconf-archive-native ncurses libxcrypt-native" DEPENDS:append:class-target = " python3-native" DEPENDS:append:class-nativesdk = " python3-native"
crypt.h is otherwise taken from the host machine Signed-off-by: Markus Volk <f_l_k@t-online.de> --- meta/recipes-devtools/python/python3_3.11.2.bb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)