[OE-core] [PATCH] qemu: IRQ Stats not shown with virsh command
Issue: LIN10-5872
Add IRQ statistic code in kvm/i8259.c.
(LOCAL REV: NOT UPSTREAM) -- oe-core already included it.
Signed-off-by: Jiping Ma
---
...1-kvm-i8259-support-info-pic-and-info-irq.patch | 76 +
.../0002-i8259-use-DEBUG_IRQ_COUNT-always.patch| 89 +++
...59-generalize-statistics-into-common-code.patch | 177 +
meta/recipes-devtools/qemu/qemu_2.10.0.bb | 3 +
4 files changed, 345 insertions(+)
create mode 100644
meta/recipes-devtools/qemu/qemu/0001-kvm-i8259-support-info-pic-and-info-irq.patch
create mode 100644
meta/recipes-devtools/qemu/qemu/0002-i8259-use-DEBUG_IRQ_COUNT-always.patch
create mode 100644
meta/recipes-devtools/qemu/qemu/0003-i8259-generalize-statistics-into-common-code.patch
diff --git
a/meta/recipes-devtools/qemu/qemu/0001-kvm-i8259-support-info-pic-and-info-irq.patch
b/meta/recipes-devtools/qemu/qemu/0001-kvm-i8259-support-info-pic-and-info-irq.patch
new file mode 100644
index 00..1640ac403d
--- /dev/null
+++
b/meta/recipes-devtools/qemu/qemu/0001-kvm-i8259-support-info-pic-and-info-irq.patch
@@ -0,0 +1,76 @@
+From ba55afb7804d9d77234440b4cf00529b052e8b87 Mon Sep 17 00:00:00 2001
+From: Peter Xu
+Date: Sun, 10 Dec 2017 14:38:18 +0800
+Subject: [PATCH 1/3] kvm-i8259: support "info pic" and "info irq"
+MIME-Version: 1.0
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: 8bit
+
+Let's leverage the i8259 common code for kvm-i8259 too.
+
+I think it's still possible that stats can lost when i8259 is in kernel
+and meanwhile when irqfd is used, e.g., by vfio or vhost devices.
+However that should be rare IMHO since they should be using MSIs mostly
+if they really want performance (that's why people use vhost and device
+assignment), and no old INTx should be used. As long as the INTx users
+are emulated in QEMU the stats will be correct.
+
+For "info pic", it should be always accurate since we fetch kvm regs
+before dump.
+
+More importantly, it's just too simple to do this now - it's only 10+
+LOC to gain this feature.
+
+Upstream-Status: Backport
+
+Signed-off-by: Peter Xu
+Message-Id: <20171210063819.14892-5-pet...@redhat.com>
+Reviewed-by: Philippe Mathieu-Daudé
+Signed-off-by: Paolo Bonzini
+Signed-off-by: Jiping Ma
+---
+ hw/i386/kvm/i8259.c | 8
+ 1 file changed, 8 insertions(+)
+
+diff --git a/hw/i386/kvm/i8259.c b/hw/i386/kvm/i8259.c
+index 11d1b726..57abe091 100644
+--- a/hw/i386/kvm/i8259.c
b/hw/i386/kvm/i8259.c
+@@ -111,6 +111,7 @@ static void kvm_pic_set_irq(void *opaque, int irq, int
level)
+ {
+ int delivered;
+
++pic_stat_update_irq(irq, level);
+ delivered = kvm_set_irq(kvm_state, irq, level);
+ apic_report_irq_delivered(delivered);
+ }
+@@ -139,12 +140,15 @@ static void kvm_i8259_class_init(ObjectClass *klass,
void *data)
+ KVMPICClass *kpc = KVM_PIC_CLASS(klass);
+ PICCommonClass *k = PIC_COMMON_CLASS(klass);
+ DeviceClass *dc = DEVICE_CLASS(klass);
++InterruptStatsProviderClass *ic = INTERRUPT_STATS_PROVIDER_CLASS(klass);
+
+ dc->reset = kvm_pic_reset;
+ kpc->parent_realize = dc->realize;
+ dc->realize = kvm_pic_realize;
+ k->pre_save = kvm_pic_get;
+ k->post_load = kvm_pic_put;
++ic->get_statistics = pic_get_statistics;
++ic->print_info = pic_print_info;
+ }
+
+ static const TypeInfo kvm_i8259_info = {
+@@ -153,6 +157,10 @@ static const TypeInfo kvm_i8259_info = {
+ .instance_size = sizeof(PICCommonState),
+ .class_init = kvm_i8259_class_init,
+ .class_size = sizeof(KVMPICClass),
++.interfaces = (InterfaceInfo[]) {
++{ TYPE_INTERRUPT_STATS_PROVIDER },
++{ }
++},
+ };
+
+ static void kvm_pic_register_types(void)
+--
+2.18.1
+
diff --git
a/meta/recipes-devtools/qemu/qemu/0002-i8259-use-DEBUG_IRQ_COUNT-always.patch
b/meta/recipes-devtools/qemu/qemu/0002-i8259-use-DEBUG_IRQ_COUNT-always.patch
new file mode 100644
index 00..f6484de67e
--- /dev/null
+++
b/meta/recipes-devtools/qemu/qemu/0002-i8259-use-DEBUG_IRQ_COUNT-always.patch
@@ -0,0 +1,89 @@
+From 67cdfed7e6515e773bd5927a1a3c2860537d4d50 Mon Sep 17 00:00:00 2001
+From: Peter Xu
+Date: Sun, 10 Dec 2017 14:38:16 +0800
+Subject: [PATCH 2/3] i8259: use DEBUG_IRQ_COUNT always
+MIME-Version: 1.0
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: 8bit
+
+It's not really scary to even enable it forever. After all it's i8259,
+and it's even not the kernel one.
+
+Then we can remove quite a few of lines to make it cleaner. And "info
+irq" will always work for it.
+
+Upstream-Status: Backport
+
+Signed-off-by: Peter Xu
+Message-Id: <20171210063819.14892-3-pet...@redhat.com>
+Reviewed-by: Philippe Mathieu-Daudé
+Signed-off-by: Paolo Bonzini
+Signed-off-by: Jiping Ma
+---
+ hw/intc/i8259.c | 17 +
+ 1 file changed, 1 insertion(+), 16 deletions(-)
+
+diff --git a/hw/intc/i8259.c b/hw/intc/i8259.c
+index fe9ecd6b..993aef9d 100644
+--- a/hw/intc/i8259.c
[OE-core] [PATCH]dhcp:"dhclient -x eth0" action is not correct.
The action of "dhclient -x eth0" and "dhclient -r eth0" is same when enable ENABLE_GENTLE_SHUTDOWN. Disable ENABLE_GENTLE_SHUTDOWN that will use the default signal hander. Signed-off-by: Jiping Ma --- meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb | 1 - 1 file changed, 1 deletion(-) diff --git a/meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb b/meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb index 19524cb3c9..275961a603 100644 --- a/meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb +++ b/meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb @@ -5,7 +5,6 @@ SRC_URI += "file://0001-define-macro-_PATH_DHCPD_CONF-and-_PATH_DHCLIENT_CON.pat file://0003-link-with-lcrypto.patch \ file://0004-Fix-out-of-tree-builds.patch \ file://0005-dhcp-client-fix-invoke-dhclient-script-failed-on-Rea.patch \ -file://0006-site.h-enable-gentle-shutdown.patch \ file://0007-Add-configure-argument-to-make-the-libxml2-dependenc.patch \ file://0009-remove-dhclient-script-bash-dependency.patch \ file://0012-dhcp-correct-the-intention-for-xml2-lib-search.patch \ -- 2.14.1 -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] possible issue with buildhistory?
hi,
not sure if it's been discussed reported before. but i ended up in a
situation where a package is "empty" (e.g. it's not created) even
though "files-in-packages.txt" in its buildhistory folder shows it
should contain a file.
I have a synthetic example:
$ cat buildhistory/packages/aarch64-linaro-linux/helloworld/latest
PV = 1.0
PR = r0
DEPENDS = autoconf-native automake-native gnu-config-native
libtool-cross libtool-native virtual/aarch64-linaro-linux-compilerlibs
virtual/aarch64-linaro-linux-gcc virtual/libc
PACKAGES = helloworld-src helloworld-dbg helloworld-staticdev
helloworld-dev helloworld-doc helloworld-locale helloworld
goodbyeworld
LAYER = meta-rpb
then:
$ cat
buildhistory/packages/aarch64-linaro-linux/helloworld/goodbyeworld/files-in-package.txt
drwxr-xr-x root root 4096 ./usr
drwxr-xr-x root root 4096 ./usr/bin
-rwxr-xr-x root root 6144 ./usr/bin/goodbyeworld
However goodbyeworld package does not exist, and /usr/bin/goodbyeworld
is in the helloworld package:
$ dpkg -c tmp-rpb-glibc/deploy/ipk/aarch64/helloworld_1.0-r0_aarch64.ipk
drwxrwxrwx root/root 0 2019-05-23 08:25 ./
drwxr-xr-x root/root 0 2019-05-23 08:25 ./usr/
drwxr-xr-x root/root 0 2019-05-23 08:25 ./usr/bin/
-rwxr-xr-x root/root 6144 2019-05-23 08:25 ./usr/bin/goodbyeworld
-rwxr-xr-x root/root 6144 2019-05-23 08:25 ./usr/bin/helloworld
It is worth saying that the file should indeed be in helloworld
package since there was a bug in my recipe! So buildhistory is wrong
here.
The recipe is a simple one:
SRC_URI = "file://helloworld.c \
file://goodbyeworld.c \
file://Makefile.am \
file://configure.ac \
file://COPYING"
S = "${WORKDIR}"
inherit autotools
PACKAGES += "goodbyeworld"
FILES_goodbyeworld += "${bindir}/goodbyeworld"
Clearly the recipe has a bug, and if I do this instead:
PACKAGES =+ "goodbyeworld"
/usr/bin/goodbyeworld ends up in the right package... however I think
in the previous case, buildhistory should have shown
/usb/bin/goodbyeworld in helloworld package, not goodbyeworld...
I was testing on warrior, but i can try to reproduce on master if needed..
thanks!
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH] libva: upgrade 2.4.0 -> 2.4.1
Changes in this release:
979a351 [common] Add A2RGB10 fourcc definition
22cba79 build: meson: enables va messaging and visibility
b89f6ec va/va_trace: add trace support for RIR(rolling intra refresh).
844b90c va/va_trace: add trace support for ROI(region of interest).
Signed-off-by: Anuj Mittal
---
.../recipes-graphics/libva/{libva_2.4.0.bb => libva_2.4.1.bb} | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
rename meta/recipes-graphics/libva/{libva_2.4.0.bb => libva_2.4.1.bb} (92%)
diff --git a/meta/recipes-graphics/libva/libva_2.4.0.bb
b/meta/recipes-graphics/libva/libva_2.4.1.bb
similarity index 92%
rename from meta/recipes-graphics/libva/libva_2.4.0.bb
rename to meta/recipes-graphics/libva/libva_2.4.1.bb
index ffa1ab8f7b..525721f0ae 100644
--- a/meta/recipes-graphics/libva/libva_2.4.0.bb
+++ b/meta/recipes-graphics/libva/libva_2.4.1.bb
@@ -19,8 +19,8 @@ LIC_FILES_CHKSUM =
"file://COPYING;md5=2e48940f94acb0af582e5ef03537800f"
SRC_URI =
"https://github.com/intel/${BPN}/releases/download/${PV}/${BP}.tar.bz2;
-SRC_URI[md5sum] = "cfc69c2a5f526dd1858e098fb8eebfa6"
-SRC_URI[sha256sum] =
"99263056c21593a26f2ece812aee6fe60142b49e6cd46cb33c8dddf18fc19391"
+SRC_URI[md5sum] = "5b5ace9de3f07cb7b8f4d19b6979adf0"
+SRC_URI[sha256sum] =
"e9e053908591b121793eaa5d8aa37675b4cd3af4b12f1f377dff4767f39cee70"
UPSTREAM_CHECK_URI = "https://github.com/intel/libva/releases;
--
2.20.1
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [V2][PATCH] dhcp:"dhclient -x eth0" action is not correct.
Please ignore V1 of [OE-core][PATCH] dhcp:"dhclient -x eth0" action is not correct. The action of "dhclient -x eth0" and "dhclient -r eth0" is same when enable ENABLE_GENTLE_SHUTDOWN. Disable ENABLE_GENTLE_SHUTDOWN that will use the default signal hander. Signed-off-by: Jiping Ma --- .../dhcp/0006-site.h-enable-gentle-shutdown.patch | 27 -- meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb | 1 - 2 files changed, 28 deletions(-) delete mode 100644 meta/recipes-connectivity/dhcp/dhcp/0006-site.h-enable-gentle-shutdown.patch diff --git a/meta/recipes-connectivity/dhcp/dhcp/0006-site.h-enable-gentle-shutdown.patch b/meta/recipes-connectivity/dhcp/dhcp/0006-site.h-enable-gentle-shutdown.patch deleted file mode 100644 index 6ef70cc..000 --- a/meta/recipes-connectivity/dhcp/dhcp/0006-site.h-enable-gentle-shutdown.patch +++ /dev/null @@ -1,27 +0,0 @@ -From 01641d146e4e6bea954e4a4ee1f6230b822665b4 Mon Sep 17 00:00:00 2001 -From: Chen Qi -Date: Tue, 15 Aug 2017 15:37:49 +0800 -Subject: [PATCH 06/11] site.h: enable gentle shutdown - -Upstream-Status: Inappropriate [configuration] -Signed-off-by: Chen Qi - -Rebase to 4.3.6 -Signed-off-by: Hongxu Jia - includes/site.h | 2 +- - 1 file changed, 1 insertion(+), 1 deletion(-) - -Index: dhcp-4.4.1/includes/site.h -=== dhcp-4.4.1.orig/includes/site.h -+++ dhcp-4.4.1/includes/site.h -@@ -295,7 +295,7 @@ -situations. We plan to revisit this feature and may -make non-backwards compatible changes including the -removal of this define. Use at your own risk. */ --/* #define ENABLE_GENTLE_SHUTDOWN */ -+#define ENABLE_GENTLE_SHUTDOWN - - /* Include old error codes. This is provided in case you -are building an external program similar to omshell for diff --git a/meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb b/meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb index ca0daa1..b4c0d5d 100644 --- a/meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb +++ b/meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb @@ -5,7 +5,6 @@ SRC_URI += "file://0001-define-macro-_PATH_DHCPD_CONF-and-_PATH_DHCLIENT_CON.pat file://0003-link-with-lcrypto.patch \ file://0004-Fix-out-of-tree-builds.patch \ file://0005-dhcp-client-fix-invoke-dhclient-script-failed-on-Rea.patch \ -file://0006-site.h-enable-gentle-shutdown.patch \ file://0007-Add-configure-argument-to-make-the-libxml2-dependenc.patch \ file://0009-remove-dhclient-script-bash-dependency.patch \ file://0012-dhcp-correct-the-intention-for-xml2-lib-search.patch \ -- 1.9.1 -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH 3/3] elfutils: Fix eu-* utils builds for musl
On Wed, May 22, 2019 at 04:29:33PM -0700, Khem Raj wrote: > On Tue, May 21, 2019 at 11:49 PM Adrian Bunk wrote: > > > > On Tue, May 21, 2019 at 06:32:28PM -0700, Khem Raj wrote: > > >... > > > PN and PN-binutils is not empty anymore on musl > > >... > > > ++#if !defined(FNM_EXTMATCH) > > > ++# define FNM_EXTMATCH (1 << 5) /* Use ksh-like extended matching. */ > > > ++#endif > > >... > > > > This might fix the compilation, but it is horribly wrong. > > > > It passes a flag to a function that this function does not understand. > > > > In the best case the function returns with an error on unknown flags. > > > > In the worst case the function does silently something different from > > what the caller expects it to do. > > > > I forgot that the API is provided by libc here so mimic'ing the define > infact could > cause pain. I think its better to nullify it in this case where it is > being used in these > files. Just looking at the usage in src/elfcompress.c, my first impression is that this tool is mostly useless without a properly working FNM_EXTMATCH. Are the eu-* utils actually important in the cases where space is so limited that people are forced to use musl instead of glibc? cu Adrian -- "Is there not promise of rain?" Ling Tan asked suddenly out of the darkness. There had been need of rain for many days. "Only a promise," Lao Er said. Pearl S. Buck - Dragon Seed -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH] [V2][PATCH] dhcp:"dhclient -x eth0" action is not correct.
Please ignore V1 of [OE-core][PATCH] dhcp:"dhclient -x eth0" action is not correct. The action of "dhclient -x eth0" and "dhclient -r eth0" is same when enable ENABLE_GENTLE_SHUTDOWN. Disable ENABLE_GENTLE_SHUTDOWN that will use the default signal hander. Signed-off-by: Jiping Ma --- .../dhcp/0006-site.h-enable-gentle-shutdown.patch | 27 -- meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb | 1 - 2 files changed, 28 deletions(-) delete mode 100644 meta/recipes-connectivity/dhcp/dhcp/0006-site.h-enable-gentle-shutdown.patch diff --git a/meta/recipes-connectivity/dhcp/dhcp/0006-site.h-enable-gentle-shutdown.patch b/meta/recipes-connectivity/dhcp/dhcp/0006-site.h-enable-gentle-shutdown.patch deleted file mode 100644 index 6ef70cc..000 --- a/meta/recipes-connectivity/dhcp/dhcp/0006-site.h-enable-gentle-shutdown.patch +++ /dev/null @@ -1,27 +0,0 @@ -From 01641d146e4e6bea954e4a4ee1f6230b822665b4 Mon Sep 17 00:00:00 2001 -From: Chen Qi -Date: Tue, 15 Aug 2017 15:37:49 +0800 -Subject: [PATCH 06/11] site.h: enable gentle shutdown - -Upstream-Status: Inappropriate [configuration] -Signed-off-by: Chen Qi - -Rebase to 4.3.6 -Signed-off-by: Hongxu Jia - includes/site.h | 2 +- - 1 file changed, 1 insertion(+), 1 deletion(-) - -Index: dhcp-4.4.1/includes/site.h -=== dhcp-4.4.1.orig/includes/site.h -+++ dhcp-4.4.1/includes/site.h -@@ -295,7 +295,7 @@ -situations. We plan to revisit this feature and may -make non-backwards compatible changes including the -removal of this define. Use at your own risk. */ --/* #define ENABLE_GENTLE_SHUTDOWN */ -+#define ENABLE_GENTLE_SHUTDOWN - - /* Include old error codes. This is provided in case you -are building an external program similar to omshell for diff --git a/meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb b/meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb index ca0daa1..b4c0d5d 100644 --- a/meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb +++ b/meta/recipes-connectivity/dhcp/dhcp_4.4.1.bb @@ -5,7 +5,6 @@ SRC_URI += "file://0001-define-macro-_PATH_DHCPD_CONF-and-_PATH_DHCLIENT_CON.pat file://0003-link-with-lcrypto.patch \ file://0004-Fix-out-of-tree-builds.patch \ file://0005-dhcp-client-fix-invoke-dhclient-script-failed-on-Rea.patch \ -file://0006-site.h-enable-gentle-shutdown.patch \ file://0007-Add-configure-argument-to-make-the-libxml2-dependenc.patch \ file://0009-remove-dhclient-script-bash-dependency.patch \ file://0012-dhcp-correct-the-intention-for-xml2-lib-search.patch \ -- 1.9.1 -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH] connman: fix segfault with musl >2.21
musl > 2.21 changed the implementation of the freeaddrinfo() function
not allowing anymore to pass null pointers to it.
This was causing a segmentation fault in connman.
---
...-gweb-fix-segfault-with-musl-v1.1.21.patch | 34 +++
.../connman/connman_1.37.bb | 5 ++-
2 files changed, 38 insertions(+), 1 deletion(-)
create mode 100644
meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
diff --git
a/meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
b/meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
new file mode 100644
index 00..43b43bc9f8
--- /dev/null
+++
b/meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
@@ -0,0 +1,34 @@
+From f0a8c69971b30ea7ca255bb885fdd1179fa5d298 Mon Sep 17 00:00:00 2001
+From: Nicola Lunghi
+Date: Thu, 23 May 2019 07:55:25 +0100
+Subject: [PATCH] gweb: fix segfault with musl v1.1.21
+
+In musl > 1.1.21 freeaddrinfo() implementation changed and
+was causing a segmentation fault on recent Yocto using musl.
+
+See this commit:
+
+
https://git.musl-libc.org/cgit/musl/commit/src/network/freeaddrinfo.c?id=d1395c43c019aec6b855cf3c656bf47c8a719e7f
+
+Upstream-Status: Submitted
+---
+ gweb/gweb.c | 3 ++-
+ 1 file changed, 2 insertions(+), 1 deletion(-)
+
+diff --git a/gweb/gweb.c b/gweb/gweb.c
+index 393afe0a..12fcb1d8 100644
+--- a/gweb/gweb.c
b/gweb/gweb.c
+@@ -1274,7 +1274,8 @@ static bool is_ip_address(const char *host)
+ addr = NULL;
+
+ result = getaddrinfo(host, NULL, , );
+- freeaddrinfo(addr);
++ if(!result)
++ freeaddrinfo(addr);
+
+ return result == 0;
+ }
+--
+2.19.1
+
diff --git a/meta/recipes-connectivity/connman/connman_1.37.bb
b/meta/recipes-connectivity/connman/connman_1.37.bb
index 2cf904cd85..f52b21cae3 100644
--- a/meta/recipes-connectivity/connman/connman_1.37.bb
+++ b/meta/recipes-connectivity/connman/connman_1.37.bb
@@ -7,7 +7,10 @@ SRC_URI =
"${KERNELORG_MIRROR}/linux/network/${BPN}/${BP}.tar.xz \
file://no-version-scripts.patch \
"
-SRC_URI_append_libc-musl = "
file://0002-resolve-musl-does-not-implement-res_ninit.patch"
+SRC_URI_append_libc-musl = " \
+file://0002-resolve-musl-does-not-implement-res_ninit.patch \
+file://0003-gweb-fix-segfault-with-musl-v1.1.21.patch \
+"
SRC_URI[md5sum] = "75012084f14fb63a84b116e66c6e94fb"
SRC_URI[sha256sum] =
"6ce29b3eb0bb16a7387bc609c39455fd13064bdcde5a4d185fab3a0c71946e16"
--
2.19.1
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH] package_manager: add help packages based on languages selected in IMAGE_LINGUAS
This is the one-line mentioned in [1]. [1] http://lists.openembedded.org/pipermail/openembedded-devel/2019-May/199769.html Signed-off-by: Andreas Müller --- meta/lib/oe/package_manager.py | 1 + 1 file changed, 1 insertion(+) diff --git a/meta/lib/oe/package_manager.py b/meta/lib/oe/package_manager.py index 2835c1ddf8..bd2432f6d3 100644 --- a/meta/lib/oe/package_manager.py +++ b/meta/lib/oe/package_manager.py @@ -566,6 +566,7 @@ class PackageManager(object, metaclass=ABCMeta): for lang in split_linguas: globs += " *-locale-%s" % lang +globs += " *-help-%s" % lang if globs is None: return -- 2.20.1 -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH 3/3] elfutils: Fix eu-* utils builds for musl
On Wed, May 22, 2019 at 11:10 PM Adrian Bunk wrote: > On Wed, May 22, 2019 at 04:29:33PM -0700, Khem Raj wrote: > > On Tue, May 21, 2019 at 11:49 PM Adrian Bunk wrote: > > > > > > On Tue, May 21, 2019 at 06:32:28PM -0700, Khem Raj wrote: > > > >... > > > > PN and PN-binutils is not empty anymore on musl > > > >... > > > > ++#if !defined(FNM_EXTMATCH) > > > > ++# define FNM_EXTMATCH (1 << 5) /* Use ksh-like extended matching. > */ > > > > ++#endif > > > >... > > > > > > This might fix the compilation, but it is horribly wrong. > > > > > > It passes a flag to a function that this function does not understand. > > > > > > In the best case the function returns with an error on unknown flags. > > > > > > In the worst case the function does silently something different from > > > what the caller expects it to do. > > > > > > > I forgot that the API is provided by libc here so mimic'ing the define > > infact could > > cause pain. I think its better to nullify it in this case where it is > > being used in these > > files. > > Just looking at the usage in src/elfcompress.c, my first impression is > that this tool is mostly useless without a properly working FNM_EXTMATCH. > > Are the eu-* utils actually important in the cases where space is so > limited that people are forced to use musl instead of glibc? > > They are primarily needed for testing and now we have good auto testing infra it would be good to build them cu > Adrian > > -- > >"Is there not promise of rain?" Ling Tan asked suddenly out > of the darkness. There had been need of rain for many days. >"Only a promise," Lao Er said. >Pearl S. Buck - Dragon Seed > > -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] librsvg: Replace workaround for old host systems with upstream fix
Just to make it clean: this flow of cleanup patches is *much
appreciated*. Thanks!
Ross
On Wed, 22 May 2019 at 11:59, Adrian Bunk wrote:
>
> Signed-off-by: Adrian Bunk
> ---
> ...t-Bsymbolic-fixes-configure-on-macOS.patch | 35 +++
> meta/recipes-gnome/librsvg/librsvg_2.40.20.bb | 8 ++---
> 2 files changed, 38 insertions(+), 5 deletions(-)
> create mode 100644
> meta/recipes-gnome/librsvg/librsvg/0001-Auto-detect-Bsymbolic-fixes-configure-on-macOS.patch
>
> diff --git
> a/meta/recipes-gnome/librsvg/librsvg/0001-Auto-detect-Bsymbolic-fixes-configure-on-macOS.patch
>
> b/meta/recipes-gnome/librsvg/librsvg/0001-Auto-detect-Bsymbolic-fixes-configure-on-macOS.patch
> new file mode 100644
> index 00..954bb60880
> --- /dev/null
> +++
> b/meta/recipes-gnome/librsvg/librsvg/0001-Auto-detect-Bsymbolic-fixes-configure-on-macOS.patch
> @@ -0,0 +1,35 @@
> +From b99891e31eb6ce550e7e1cb2ca592095b3050a93 Mon Sep 17 00:00:00 2001
> +From: Brion Vibber
> +Date: Sun, 25 Feb 2018 18:42:36 -0800
> +Subject: Auto-detect -Bsymbolic, fixes configure on macOS
> +
> +The -Bsymbolic linker option is ELF-specific, and was breaking
> +configure on macOS unless --disable-Bsymbolic was explicitly passed.
> +
> +Switching the behavior from requiring -Bsymbolic to be available
> +by default to just warning and continuing on without.
> +
> +Fixes https://gitlab.gnome.org/GNOME/librsvg/issues/211
> +
> +Upstream-Status: Backport
> +Signed-off-by: Adrian Bunk
> +---
> + configure.ac | 2 +-
> + 1 file changed, 1 insertion(+), 1 deletion(-)
> +
> +diff --git a/configure.ac b/configure.ac
> +index 15b26b2d..9f8dce29 100644
> +--- a/configure.ac
> b/configure.ac
> +@@ -216,7 +216,7 @@ AM_CONDITIONAL([ENABLE_PIXBUF_LOADER],[test
> "$enable_pixbuf_loader" = "yes"])
> + AC_ARG_ENABLE([Bsymbolic],
> + [AS_HELP_STRING([--disable-Bsymbolic],
> + [disable linking with -Bsymbolic])],
> +- [],[enable_Bsymbolic=yes])
> ++ [enable_Bsymbolic=no],[enable_Bsymbolic=auto])
> +
> + BSYMBOLIC_LDFLAG=
> + if test "$enable_Bsymbolic" != "no"; then
> +--
> +2.20.1
> +
> diff --git a/meta/recipes-gnome/librsvg/librsvg_2.40.20.bb
> b/meta/recipes-gnome/librsvg/librsvg_2.40.20.bb
> index e376d52ef3..6a798e6a91 100644
> --- a/meta/recipes-gnome/librsvg/librsvg_2.40.20.bb
> +++ b/meta/recipes-gnome/librsvg/librsvg_2.40.20.bb
> @@ -14,17 +14,15 @@ BBCLASSEXTEND = "native"
>
> inherit gnomebase gtk-doc pixbufcache upstream-version-is-even
> gobject-introspection
>
> -SRC_URI += "file://gtk-option.patch"
> +SRC_URI += "file://gtk-option.patch \
> +file://0001-Auto-detect-Bsymbolic-fixes-configure-on-macOS.patch
> \
> +"
>
> SRC_URI[archive.md5sum] = "4949d313b0c5d9161a5c259104af5568"
> SRC_URI[archive.sha256sum] =
> "cff4dd3c3b78bfe99d8fcfad3b8ba1eee3289a0823c0e118d78106be6b84c92b"
>
> CACHED_CONFIGUREVARS =
> "ac_cv_path_GDK_PIXBUF_QUERYLOADERS=${STAGING_LIBDIR_NATIVE}/gdk-pixbuf-2.0/gdk-pixbuf-query-loaders"
>
> -# The older ld (2.22) on the host (Centos 6.5) doesn't have the
> -# -Bsymbolic-functions option, we can disable it for native.
> -EXTRA_OECONF_append_class-native = " --enable-Bsymbolic=auto"
> -
> PACKAGECONFIG ??= "gdkpixbuf"
> # The gdk-pixbuf loader
> PACKAGECONFIG[gdkpixbuf] =
> "--enable-pixbuf-loader,--disable-pixbuf-loader,gdk-pixbuf-native"
> --
> 2.17.1
>
> --
> ___
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] package_manager: add help packages based on languages selected in IMAGE_LINGUAS
Why -help- and not -doc? Ross On Thu, 23 May 2019 at 13:30, Andreas Müller wrote: > > This is the one-line mentioned in [1]. > > [1] > http://lists.openembedded.org/pipermail/openembedded-devel/2019-May/199769.html > > Signed-off-by: Andreas Müller > --- > meta/lib/oe/package_manager.py | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/meta/lib/oe/package_manager.py b/meta/lib/oe/package_manager.py > index 2835c1ddf8..bd2432f6d3 100644 > --- a/meta/lib/oe/package_manager.py > +++ b/meta/lib/oe/package_manager.py > @@ -566,6 +566,7 @@ class PackageManager(object, metaclass=ABCMeta): > > for lang in split_linguas: > globs += " *-locale-%s" % lang > +globs += " *-help-%s" % lang > > if globs is None: > return > -- > 2.20.1 > > -- > ___ > Openembedded-core mailing list > Openembedded-core@lists.openembedded.org > http://lists.openembedded.org/mailman/listinfo/openembedded-core -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] package_manager: add help packages based on languages selected in IMAGE_LINGUAS
On Thu, 23 May 2019 at 13:44, Andreas Müller wrote: > > Why -help- and not -doc? > > > Because there are help files installed in 'help' directory and > displayed by a help browser. Sounds a lot like documentation to me... -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH] autotools, cmake, meson, waf: define PACKAGECONFIG_CONFARGS before appending it
* just to make sure it's expaned by bitbake before it gets executed in shell
* e.g. with cmake.bbclass and cmake recipe (any recipe without
PACKAGECONFIG options have this issue) it looks like this:
bitbake -e cmake | grep EXTRA_OECMAKE=
EXTRA_OECMAKE=" -DCMAKE_DOC_DIR=share/doc/cmake-3.14
-DCMAKE_USE_SYSTEM_LIBRARIES=1 -DCMAKE_USE_SYSTEM_LIBRARY_JSONCPP=0
-DCMAKE_USE_SYSTEM_LIBRARY_LIBUV=0
-DCMAKE_USE_SYSTEM_LIBRARY_LIBRHASH=0 -DKWSYS_CHAR_IS_SIGNED=1
-DBUILD_CursesDialog=0 -DKWSYS_LFS_WORKS=1
\${PACKAGECONFIG_CONFARGS}"
* there are some other places where PACKAGECONFIG_CONFARGS is used, but
looks like all of them started to use it only after adding
PACKAGECONFIG options in the recipe
Signed-off-by: Martin Jansa
---
meta/classes/autotools.bbclass | 1 +
meta/classes/cmake.bbclass | 1 +
meta/classes/meson.bbclass | 1 +
meta/classes/waf.bbclass | 1 +
4 files changed, 4 insertions(+)
diff --git a/meta/classes/autotools.bbclass b/meta/classes/autotools.bbclass
index 8768a6ad68..32961320a5 100644
--- a/meta/classes/autotools.bbclass
+++ b/meta/classes/autotools.bbclass
@@ -129,6 +129,7 @@ autotools_postconfigure(){
EXTRACONFFUNCS ??= ""
+PACKAGECONFIG_CONFARGS ??= ""
EXTRA_OECONF_append = " ${PACKAGECONFIG_CONFARGS}"
do_configure[prefuncs] += "autotools_preconfigure autotools_aclocals
${EXTRACONFFUNCS}"
diff --git a/meta/classes/cmake.bbclass b/meta/classes/cmake.bbclass
index d3f0d70847..b5deb7da70 100644
--- a/meta/classes/cmake.bbclass
+++ b/meta/classes/cmake.bbclass
@@ -56,6 +56,7 @@ OECMAKE_EXTRA_ROOT_PATH ?= ""
OECMAKE_FIND_ROOT_PATH_MODE_PROGRAM = "ONLY"
OECMAKE_FIND_ROOT_PATH_MODE_PROGRAM_class-native = "BOTH"
+PACKAGECONFIG_CONFARGS ??= ""
EXTRA_OECMAKE_append = " ${PACKAGECONFIG_CONFARGS}"
EXTRA_OECMAKE_BUILD_prepend_task-compile = "${PARALLEL_MAKE} "
diff --git a/meta/classes/meson.bbclass b/meta/classes/meson.bbclass
index 115d1aedcb..4f03a51e09 100644
--- a/meta/classes/meson.bbclass
+++ b/meta/classes/meson.bbclass
@@ -35,6 +35,7 @@ MESON_C_ARGS = "${MESON_TOOLCHAIN_ARGS} ${CFLAGS}"
MESON_CPP_ARGS = "${MESON_TOOLCHAIN_ARGS} ${CXXFLAGS}"
MESON_LINK_ARGS = "${MESON_TOOLCHAIN_ARGS} ${LDFLAGS}"
+PACKAGECONFIG_CONFARGS ??= ""
EXTRA_OEMESON_append = " ${PACKAGECONFIG_CONFARGS}"
MESON_CROSS_FILE = ""
diff --git a/meta/classes/waf.bbclass b/meta/classes/waf.bbclass
index 900244004e..f7178b28e9 100644
--- a/meta/classes/waf.bbclass
+++ b/meta/classes/waf.bbclass
@@ -3,6 +3,7 @@ DISABLE_STATIC = ""
B = "${WORKDIR}/build"
+PACKAGECONFIG_CONFARGS ??= ""
EXTRA_OECONF_append = " ${PACKAGECONFIG_CONFARGS}"
def waflock_hash(d):
--
2.17.1
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] connman: fix segfault with musl >2.21
On Thu, 2019-05-23 at 12:06 +0100, nick83ola wrote:
> Hi Andre,
>
> in glibc freeaddrinfo is checking for null
>
> https://github.molgen.mpg.de/git-mirror/glibc/blob/20003c49884422da7ffbc459cdeee768a6fee07b/sysdeps/posix/getaddrinfo.c#L2663-L2674
>
> void
> freeaddrinfo (struct addrinfo *ai)
> {
> struct addrinfo *p;
>
> while (ai != NULL)
> {
> p = ai;
> ai = ai->ai_next;
> free (p->ai_canonname);
> free (p);
> }
> }
This is an implementation detail that isn't mandated by the standard (and
hence could change).
Quoting
https://pubs.opengroup.org/onlinepubs/9699919799/functions/freeaddrinfo.html
'The freeaddrinfo() function shall free one or more addrinfo structures
returned by getaddrinfo(), along with any additional storage associated
with those structures'
Technically, unless getaddrinfo() succeeds, one didn't get an
'addrinfo structure returned by getaddrinfo()', so one equally
can't pass it into freeaddrinfo() unless getaddrinfo() succeeded.
Also, in connman's gweb.c 'addr' is initialised to NULL. Again,
NULL isn't anything 'returned by getaddrinfo()', so even in glibc
it only works by pure luck.
In other words, applying the patch in both glibc & musl environments
is the right thing to do.
See below.
>
> that is why I put it into the musl specific patches.
> But ok I will repost it putting it in the generic SRC_URI.
>
> regarding the 2.21 -> yes is a typo.
>
> Thanks
> Nicola Lunghi
>
>
> On Thu, 23 May 2019 at 10:18, André Draszik wrote:
> > On Thu, 2019-05-23 at 08:47 +0100, Nicola Lunghi wrote:
> > > musl > 2.21 changed the implementation of the freeaddrinfo() function
> > > not allowing anymore to pass null pointers to it.
> > > This was causing a segmentation fault in connman.
> > > ---
> > > ...-gweb-fix-segfault-with-musl-v1.1.21.patch | 34 +++
> > > .../connman/connman_1.37.bb | 5 ++-
> > > 2 files changed, 38 insertions(+), 1 deletion(-)
> > > create mode 100644
> > > meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
> > >
> > > diff --git
> > > a/meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
> > > b/meta/recipes-
> > > connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
> > > new file mode 100644
> > > index 00..43b43bc9f8
> > > --- /dev/null
> > > +++
> > > b/meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
> > > @@ -0,0 +1,34 @@
> > > +From f0a8c69971b30ea7ca255bb885fdd1179fa5d298 Mon Sep 17 00:00:00 2001
> > > +From: Nicola Lunghi
> > > +Date: Thu, 23 May 2019 07:55:25 +0100
> > > +Subject: [PATCH] gweb: fix segfault with musl v1.1.21
> > > +
> > > +In musl > 1.1.21 freeaddrinfo() implementation changed and
> > > +was causing a segmentation fault on recent Yocto using musl.
> > > +
> > > +See this commit:
> > > +
> > > +
> > > https://git.musl-libc.org/cgit/musl/commit/src/network/freeaddrinfo.c?id=d1395c43c019aec6b855cf3c656bf47c8a719e7f
> > > +
> > > +Upstream-Status: Submitted
> > > +---
> > > + gweb/gweb.c | 3 ++-
> > > + 1 file changed, 2 insertions(+), 1 deletion(-)
> > > +
> > > +diff --git a/gweb/gweb.c b/gweb/gweb.c
> > > +index 393afe0a..12fcb1d8 100644
> > > +--- a/gweb/gweb.c
> > > b/gweb/gweb.c
> > > +@@ -1274,7 +1274,8 @@ static bool is_ip_address(const char *host)
> > > + addr = NULL;
Ideally, this patch should also remove the assignment of addr = NULL;
as per above. But that's a discussion for the connman list.
Cheers,
Andre'
> > > +
> > > + result = getaddrinfo(host, NULL, , );
> > > +-freeaddrinfo(addr);
> > > ++if(!result)
> > > ++freeaddrinfo(addr);
> > > +
> > > + return result == 0;
> > > + }
> > > +--
> > > +2.19.1
> > > +
> > > diff --git a/meta/recipes-connectivity/connman/connman_1.37.bb
> > > b/meta/recipes-connectivity/connman/connman_1.37.bb
> > > index 2cf904cd85..f52b21cae3 100644
> > > --- a/meta/recipes-connectivity/connman/connman_1.37.bb
> > > +++ b/meta/recipes-connectivity/connman/connman_1.37.bb
> > > @@ -7,7 +7,10 @@ SRC_URI =
> > > "${KERNELORG_MIRROR}/linux/network/${BPN}/${BP}.tar.xz \
> > > file://no-version-scripts.patch \
> > > "
> > >
> > > -SRC_URI_append_libc-musl = "
> > > file://0002-resolve-musl-does-not-implement-res_ninit.patch"
> > > +SRC_URI_append_libc-musl = " \
> > > +file://0002-resolve-musl-does-not-implement-res_ninit.patch \
> > > +file://0003-gweb-fix-segfault-with-musl-v1.1.21.patch \
> > > +"
> >
> > This fix is not musl specific.
> >
> > Cheers,
> > Andre'
> >
> > > SRC_URI[md5sum] = "75012084f14fb63a84b116e66c6e94fb"
> > > SRC_URI[sha256sum] =
> > > "6ce29b3eb0bb16a7387bc609c39455fd13064bdcde5a4d185fab3a0c71946e16"
> > > --
> > > 2.19.1
> > >
> >
> > --
> > ___
> > Openembedded-core mailing list
> >
Re: [OE-core] [PATCH] autotools, cmake, meson, waf: define PACKAGECONFIG_CONFARGS before appending it
I think a neater solution might be to just aways set that when
handling PACKAGECONFIG in base.bbclass. Presumably the problem here
is that if PACKAGECONFIG isn't set then PACKAGECONFIG_CONFARGS is
never set, so lets ensure it is always set.
Ross
On Thu, 23 May 2019 at 14:55, Martin Jansa wrote:
>
> * just to make sure it's expaned by bitbake before it gets executed in shell
> * e.g. with cmake.bbclass and cmake recipe (any recipe without
> PACKAGECONFIG options have this issue) it looks like this:
> bitbake -e cmake | grep EXTRA_OECMAKE=
> EXTRA_OECMAKE=" -DCMAKE_DOC_DIR=share/doc/cmake-3.14
> -DCMAKE_USE_SYSTEM_LIBRARIES=1 -DCMAKE_USE_SYSTEM_LIBRARY_JSONCPP=0
> -DCMAKE_USE_SYSTEM_LIBRARY_LIBUV=0
> -DCMAKE_USE_SYSTEM_LIBRARY_LIBRHASH=0 -DKWSYS_CHAR_IS_SIGNED=1
> -DBUILD_CursesDialog=0 -DKWSYS_LFS_WORKS=1
> \${PACKAGECONFIG_CONFARGS}"
>
> * there are some other places where PACKAGECONFIG_CONFARGS is used, but
> looks like all of them started to use it only after adding
> PACKAGECONFIG options in the recipe
>
> Signed-off-by: Martin Jansa
> ---
> meta/classes/autotools.bbclass | 1 +
> meta/classes/cmake.bbclass | 1 +
> meta/classes/meson.bbclass | 1 +
> meta/classes/waf.bbclass | 1 +
> 4 files changed, 4 insertions(+)
>
> diff --git a/meta/classes/autotools.bbclass b/meta/classes/autotools.bbclass
> index 8768a6ad68..32961320a5 100644
> --- a/meta/classes/autotools.bbclass
> +++ b/meta/classes/autotools.bbclass
> @@ -129,6 +129,7 @@ autotools_postconfigure(){
>
> EXTRACONFFUNCS ??= ""
>
> +PACKAGECONFIG_CONFARGS ??= ""
> EXTRA_OECONF_append = " ${PACKAGECONFIG_CONFARGS}"
>
> do_configure[prefuncs] += "autotools_preconfigure autotools_aclocals
> ${EXTRACONFFUNCS}"
> diff --git a/meta/classes/cmake.bbclass b/meta/classes/cmake.bbclass
> index d3f0d70847..b5deb7da70 100644
> --- a/meta/classes/cmake.bbclass
> +++ b/meta/classes/cmake.bbclass
> @@ -56,6 +56,7 @@ OECMAKE_EXTRA_ROOT_PATH ?= ""
> OECMAKE_FIND_ROOT_PATH_MODE_PROGRAM = "ONLY"
> OECMAKE_FIND_ROOT_PATH_MODE_PROGRAM_class-native = "BOTH"
>
> +PACKAGECONFIG_CONFARGS ??= ""
> EXTRA_OECMAKE_append = " ${PACKAGECONFIG_CONFARGS}"
>
> EXTRA_OECMAKE_BUILD_prepend_task-compile = "${PARALLEL_MAKE} "
> diff --git a/meta/classes/meson.bbclass b/meta/classes/meson.bbclass
> index 115d1aedcb..4f03a51e09 100644
> --- a/meta/classes/meson.bbclass
> +++ b/meta/classes/meson.bbclass
> @@ -35,6 +35,7 @@ MESON_C_ARGS = "${MESON_TOOLCHAIN_ARGS} ${CFLAGS}"
> MESON_CPP_ARGS = "${MESON_TOOLCHAIN_ARGS} ${CXXFLAGS}"
> MESON_LINK_ARGS = "${MESON_TOOLCHAIN_ARGS} ${LDFLAGS}"
>
> +PACKAGECONFIG_CONFARGS ??= ""
> EXTRA_OEMESON_append = " ${PACKAGECONFIG_CONFARGS}"
>
> MESON_CROSS_FILE = ""
> diff --git a/meta/classes/waf.bbclass b/meta/classes/waf.bbclass
> index 900244004e..f7178b28e9 100644
> --- a/meta/classes/waf.bbclass
> +++ b/meta/classes/waf.bbclass
> @@ -3,6 +3,7 @@ DISABLE_STATIC = ""
>
> B = "${WORKDIR}/build"
>
> +PACKAGECONFIG_CONFARGS ??= ""
> EXTRA_OECONF_append = " ${PACKAGECONFIG_CONFARGS}"
>
> def waflock_hash(d):
> --
> 2.17.1
>
> --
> ___
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH] Revert "musl: Add TEMP_FAILURE_RETRY from glibc"
This reverts commit 0c68c621d3b8690e3a346170006f75b032511a8a.
It should be upstreamed first
Secondly, musl and glibc are under different licenses and importing
code from one to other could taint it.
Signed-off-by: Khem Raj
---
...0001-unistd.h-Add-TEMP_FAILURE_RETRY.patch | 36 ---
meta/recipes-core/musl/musl_git.bb| 1 -
2 files changed, 37 deletions(-)
delete mode 100644
meta/recipes-core/musl/musl/0001-unistd.h-Add-TEMP_FAILURE_RETRY.patch
diff --git
a/meta/recipes-core/musl/musl/0001-unistd.h-Add-TEMP_FAILURE_RETRY.patch
b/meta/recipes-core/musl/musl/0001-unistd.h-Add-TEMP_FAILURE_RETRY.patch
deleted file mode 100644
index 4d48d61651..00
--- a/meta/recipes-core/musl/musl/0001-unistd.h-Add-TEMP_FAILURE_RETRY.patch
+++ /dev/null
@@ -1,36 +0,0 @@
-From 19cbebc3fa33c6a1a71a6036da4d67c98f859f06 Mon Sep 17 00:00:00 2001
-From: Adrian Bunk
-Date: Wed, 15 May 2019 16:08:33 +0300
-Subject: unistd.h: Add TEMP_FAILURE_RETRY
-
-Signed-off-by: Adrian Bunk
-Upstream-Status: Inappropriate [oe-specific]
- include/unistd.h | 11 +++
- 1 file changed, 11 insertions(+)
-
-diff --git a/include/unistd.h b/include/unistd.h
-index 9485da7a..8a75af57 100644
a/include/unistd.h
-+++ b/include/unistd.h
-@@ -460,6 +460,17 @@ int eaccess(const char *, int);
- #define _CS_V6_ENV1148
- #define _CS_V7_ENV1149
-
-+#ifdef _GNU_SOURCE
-+#ifndef TEMP_FAILURE_RETRY
-+# define TEMP_FAILURE_RETRY(expression) \
-+ (__extension__
\
-+({ long int __result;
\
-+ do __result = (long int) (expression);
\
-+ while (__result == -1L && errno == EINTR);
\
-+ __result; }))
-+#endif
-+#endif
-+
- #ifdef __cplusplus
- }
- #endif
---
-2.20.1
-
diff --git a/meta/recipes-core/musl/musl_git.bb
b/meta/recipes-core/musl/musl_git.bb
index 2b433aed8c..1e324804e1 100644
--- a/meta/recipes-core/musl/musl_git.bb
+++ b/meta/recipes-core/musl/musl_git.bb
@@ -15,7 +15,6 @@ PV = "${BASEVER}+git${SRCPV}"
SRC_URI = "git://git.musl-libc.org/musl \
file://0001-Make-dynamic-linker-a-relative-symlink-to-libc.patch \
file://0002-ldso-Use-syslibdir-and-libdir-as-default-pathes-to-l.patch \
- file://0001-unistd.h-Add-TEMP_FAILURE_RETRY.patch \
"
S = "${WORKDIR}/git"
--
2.21.0
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] package_manager: add help packages based on languages selected in IMAGE_LINGUAS
On Thu, May 23, 2019 at 2:39 PM Burton, Ross wrote: > > Why -help- and not -doc? > Because there are help files installed in 'help' directory and displayed by a help browser. Andreas -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] Should systemd be marked as incompatible with musl?
On Thu, 23 May 2019 at 11:34, Adrian Bunk wrote: > systemd is fundamentally Linux-only and not portable to other kernels. > > systemd upstream is using glibc extensions not present in other > C libraries. > > systemd upstream is accepting technically correct patches that help > building with musl, but there is no interest upstream in keeping systemd > working with non-glibc C libraries. > > The way things currently go, systemd/musl will require an ever-growing > amount of not upstreamable patches - and this is not sustainable. I think I have to agree with you: several of the extensions are for security purposes, so we're potentially actively introducing issues. Ross -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] connman: fix segfault with musl >2.21
On Thu, 23 May 2019 at 12:07, nick83ola wrote: > that is why I put it into the musl specific patches. > But ok I will repost it putting it in the generic SRC_URI. Rule of thumb is to only apply patches in an override if they break other cases outside of the override. In this case, it's just an extra guard, so nothing else breaks. Rationale being that when a patch is in an override it doesn't get tested as much. Ross -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH 1/2] piglit: upgrade to latest revision
Devtool does not create commits; it's up to the developer and/or AUH.
I wouldn't necessarily want AUH to try to auto-explain the changes, as
it's worse than no explanation imo :)
Alex
On Thu, 23 May 2019 at 14:23, Burton, Ross wrote:
>
> Can you file a bug to get devtool to state that in the commit message?
>
> Ross
>
> On Thu, 23 May 2019 at 10:27, Alexander Kanavin
> wrote:
> >
> > Yes. If devtool drops them, then they are already in the code.
> >
> > Alex
> >
> > On Thu, 23 May 2019 at 03:30, Khem Raj wrote:
> > >
> > > On Wed, May 22, 2019 at 4:02 AM Alexander Kanavin
> > > wrote:
> > > >
> > > > Signed-off-by: Alexander Kanavin
> > > > ---
> > > > ...UPWARD-only-if-its-defined-in-fenv.h.patch | 75 ---
> > > > .../piglit/piglit/format-fix.patch| 69 -
> > >
> > > are these patches upstreamed ?
> > >
> > > > meta/recipes-graphics/piglit/piglit_git.bb| 5 +-
> > > > 3 files changed, 1 insertion(+), 148 deletions(-)
> > > > delete mode 100644
> > > > meta/recipes-graphics/piglit/piglit/0001-tests-Use-FE_UPWARD-only-if-its-defined-in-fenv.h.patch
> > > > delete mode 100644 meta/recipes-graphics/piglit/piglit/format-fix.patch
> > > >
> > > > diff --git
> > > > a/meta/recipes-graphics/piglit/piglit/0001-tests-Use-FE_UPWARD-only-if-its-defined-in-fenv.h.patch
> > > >
> > > > b/meta/recipes-graphics/piglit/piglit/0001-tests-Use-FE_UPWARD-only-if-its-defined-in-fenv.h.patch
> > > > deleted file mode 100644
> > > > index 57eda2e207b..000
> > > > ---
> > > > a/meta/recipes-graphics/piglit/piglit/0001-tests-Use-FE_UPWARD-only-if-its-defined-in-fenv.h.patch
> > > > +++ /dev/null
> > > > @@ -1,75 +0,0 @@
> > > > -Upstream-Status: Submitted [mailing list]
> > > > -Signed-off-by: Ross Burton
> > > > -
> > > > -From 0e0a2a69261031d55d52b6045990e8982ea12912 Mon Sep 17 00:00:00 2001
> > > > -From: Khem Raj
> > > > -Date: Sat, 9 Jul 2016 07:52:19 +
> > > > -Subject: [PATCH] tests: only run rounding tests if FE_UPWARD is present
> > > > -
> > > > -On ARM, musl does not define FE_* when the architecture does not have
> > > > VFP (which
> > > > -is the right interpretation).
> > > > -
> > > > -As these tests depend on calling fesetround(), skip the test if
> > > > FE_UPWARD isn't
> > > > -available.
> > > > -
> > > > -Signed-off-by: Ross Burton
> > > >
> > > > - tests/general/roundmode-getintegerv.c | 12
> > > > - tests/general/roundmode-pixelstore.c | 12
> > > > - 2 files changed, 16 insertions(+), 8 deletions(-)
> > > > -
> > > > -diff --git a/tests/general/roundmode-getintegerv.c
> > > > b/tests/general/roundmode-getintegerv.c
> > > > -index 28ecfaf55..aa99044a1 100644
> > > > a/tests/general/roundmode-getintegerv.c
> > > > -+++ b/tests/general/roundmode-getintegerv.c
> > > > -@@ -79,13 +79,17 @@ test(float val, int expect)
> > > > - void
> > > > - piglit_init(int argc, char **argv)
> > > > - {
> > > > -- int ret;
> > > > - bool pass = true;
> > > > -- ret = fesetround(FE_UPWARD);
> > > > -- if (ret != 0) {
> > > > -- printf("Couldn't set rounding mode\n");
> > > > -+
> > > > -+#ifdef FE_UPWARD
> > > > -+ if (fesetround(FE_UPWARD) != 0) {
> > > > -+ printf("Setting rounding mode failed\n");
> > > > - piglit_report_result(PIGLIT_SKIP);
> > > > - }
> > > > -+#else
> > > > -+ printf("Cannot set rounding mode\n");
> > > > -+ piglit_report_result(PIGLIT_SKIP);
> > > > -+#endif
> > > > -
> > > > - pass = test(2.2, 2) && pass;
> > > > - pass = test(2.8, 3) && pass;
> > > > -diff --git a/tests/general/roundmode-pixelstore.c
> > > > b/tests/general/roundmode-pixelstore.c
> > > > -index 8a029b257..57ec11c09 100644
> > > > a/tests/general/roundmode-pixelstore.c
> > > > -+++ b/tests/general/roundmode-pixelstore.c
> > > > -@@ -79,13 +79,17 @@ test(float val, int expect)
> > > > - void
> > > > - piglit_init(int argc, char **argv)
> > > > - {
> > > > -- int ret;
> > > > - bool pass = true;
> > > > -- ret = fesetround(FE_UPWARD);
> > > > -- if (ret != 0) {
> > > > -- printf("Couldn't set rounding mode\n");
> > > > -+
> > > > -+#ifdef FE_UPWARD
> > > > -+ if (fesetround(FE_UPWARD) != 0) {
> > > > -+ printf("Setting rounding mode failed\n");
> > > > - piglit_report_result(PIGLIT_SKIP);
> > > > - }
> > > > -+#else
> > > > -+ printf("Cannot set rounding mode\n");
> > > > -+ piglit_report_result(PIGLIT_SKIP);
> > > > -+#endif
> > > > -
> > > > - pass = test(2.2, 2) && pass;
> > > > - pass = test(2.8, 3) && pass;
> > > > ---
> > > > -2.11.0
> > > > -
> > > > diff --git a/meta/recipes-graphics/piglit/piglit/format-fix.patch
> > > > b/meta/recipes-graphics/piglit/piglit/format-fix.patch
> > > > deleted file mode 100644
> > > > index 73d539fef27..000
> > > > --- a/meta/recipes-graphics/piglit/piglit/format-fix.patch
> > > > +++
[OE-core] [PATCH v2] connman: fix segfault with musl >v1.1.21
musl > v1.1.21 changed the implementation of the freeaddrinfo() function
not allowing anymore to pass null pointers to it.
This was causing a segmentation fault in connman.
Signed-off-by: Nicola Lunghi
---
...-gweb-fix-segfault-with-musl-v1.1.21.patch | 34 +++
.../connman/connman_1.37.bb | 1 +
2 files changed, 35 insertions(+)
create mode 100644
meta/recipes-connectivity/connman/connman/0001-gweb-fix-segfault-with-musl-v1.1.21.patch
diff --git
a/meta/recipes-connectivity/connman/connman/0001-gweb-fix-segfault-with-musl-v1.1.21.patch
b/meta/recipes-connectivity/connman/connman/0001-gweb-fix-segfault-with-musl-v1.1.21.patch
new file mode 100644
index 00..43b43bc9f8
--- /dev/null
+++
b/meta/recipes-connectivity/connman/connman/0001-gweb-fix-segfault-with-musl-v1.1.21.patch
@@ -0,0 +1,34 @@
+From f0a8c69971b30ea7ca255bb885fdd1179fa5d298 Mon Sep 17 00:00:00 2001
+From: Nicola Lunghi
+Date: Thu, 23 May 2019 07:55:25 +0100
+Subject: [PATCH] gweb: fix segfault with musl v1.1.21
+
+In musl > v1.1.21 freeaddrinfo() implementation changed and
+was causing a segmentation fault on recent Yocto using musl.
+
+See this commit:
+
+
https://git.musl-libc.org/cgit/musl/commit/src/network/freeaddrinfo.c?id=d1395c43c019aec6b855cf3c656bf47c8a719e7f
+
+Upstream-Status: Submitted
+---
+ gweb/gweb.c | 3 ++-
+ 1 file changed, 2 insertions(+), 1 deletion(-)
+
+diff --git a/gweb/gweb.c b/gweb/gweb.c
+index 393afe0a..12fcb1d8 100644
+--- a/gweb/gweb.c
b/gweb/gweb.c
+@@ -1274,7 +1274,8 @@ static bool is_ip_address(const char *host)
+ addr = NULL;
+
+ result = getaddrinfo(host, NULL, , );
+- freeaddrinfo(addr);
++ if(!result)
++ freeaddrinfo(addr);
+
+ return result == 0;
+ }
+--
+2.19.1
+
diff --git a/meta/recipes-connectivity/connman/connman_1.37.bb
b/meta/recipes-connectivity/connman/connman_1.37.bb
index 2cf904cd85..12650329c5 100644
--- a/meta/recipes-connectivity/connman/connman_1.37.bb
+++ b/meta/recipes-connectivity/connman/connman_1.37.bb
@@ -3,6 +3,7 @@ require connman.inc
SRC_URI = "${KERNELORG_MIRROR}/linux/network/${BPN}/${BP}.tar.xz \
file://0001-plugin.h-Change-visibility-to-default-for-debug-symb.patch \
file://0001-connman.service-stop-systemd-resolved-when-we-use-co.patch \
+file://0001-gweb-fix-segfault-with-musl-v1.1.21.patch \
file://connman \
file://no-version-scripts.patch \
"
--
2.20.1
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] connman: fix segfault with musl >2.21
Hi Andre,
in glibc freeaddrinfo is checking for null
https://github.molgen.mpg.de/git-mirror/glibc/blob/20003c49884422da7ffbc459cdeee768a6fee07b/sysdeps/posix/getaddrinfo.c#L2663-L2674
void
freeaddrinfo (struct addrinfo *ai)
{
struct addrinfo *p;
while (ai != NULL)
{
p = ai;
ai = ai->ai_next;
free (p->ai_canonname);
free (p);
}
}
that is why I put it into the musl specific patches.
But ok I will repost it putting it in the generic SRC_URI.
regarding the 2.21 -> yes is a typo.
Thanks
Nicola Lunghi
On Thu, 23 May 2019 at 10:18, André Draszik wrote:
>
> On Thu, 2019-05-23 at 08:47 +0100, Nicola Lunghi wrote:
> > musl > 2.21 changed the implementation of the freeaddrinfo() function
> > not allowing anymore to pass null pointers to it.
> > This was causing a segmentation fault in connman.
> > ---
> > ...-gweb-fix-segfault-with-musl-v1.1.21.patch | 34 +++
> > .../connman/connman_1.37.bb | 5 ++-
> > 2 files changed, 38 insertions(+), 1 deletion(-)
> > create mode 100644
> > meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
> >
> > diff --git
> > a/meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
> > b/meta/recipes-
> > connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
> > new file mode 100644
> > index 00..43b43bc9f8
> > --- /dev/null
> > +++
> > b/meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
> > @@ -0,0 +1,34 @@
> > +From f0a8c69971b30ea7ca255bb885fdd1179fa5d298 Mon Sep 17 00:00:00 2001
> > +From: Nicola Lunghi
> > +Date: Thu, 23 May 2019 07:55:25 +0100
> > +Subject: [PATCH] gweb: fix segfault with musl v1.1.21
> > +
> > +In musl > 1.1.21 freeaddrinfo() implementation changed and
> > +was causing a segmentation fault on recent Yocto using musl.
> > +
> > +See this commit:
> > +
> > +
> > https://git.musl-libc.org/cgit/musl/commit/src/network/freeaddrinfo.c?id=d1395c43c019aec6b855cf3c656bf47c8a719e7f
> > +
> > +Upstream-Status: Submitted
> > +---
> > + gweb/gweb.c | 3 ++-
> > + 1 file changed, 2 insertions(+), 1 deletion(-)
> > +
> > +diff --git a/gweb/gweb.c b/gweb/gweb.c
> > +index 393afe0a..12fcb1d8 100644
> > +--- a/gweb/gweb.c
> > b/gweb/gweb.c
> > +@@ -1274,7 +1274,8 @@ static bool is_ip_address(const char *host)
> > + addr = NULL;
> > +
> > + result = getaddrinfo(host, NULL, , );
> > +-freeaddrinfo(addr);
> > ++if(!result)
> > ++freeaddrinfo(addr);
> > +
> > + return result == 0;
> > + }
> > +--
> > +2.19.1
> > +
> > diff --git a/meta/recipes-connectivity/connman/connman_1.37.bb
> > b/meta/recipes-connectivity/connman/connman_1.37.bb
> > index 2cf904cd85..f52b21cae3 100644
> > --- a/meta/recipes-connectivity/connman/connman_1.37.bb
> > +++ b/meta/recipes-connectivity/connman/connman_1.37.bb
> > @@ -7,7 +7,10 @@ SRC_URI =
> > "${KERNELORG_MIRROR}/linux/network/${BPN}/${BP}.tar.xz \
> > file://no-version-scripts.patch \
> > "
> >
> > -SRC_URI_append_libc-musl = "
> > file://0002-resolve-musl-does-not-implement-res_ninit.patch"
> > +SRC_URI_append_libc-musl = " \
> > +file://0002-resolve-musl-does-not-implement-res_ninit.patch \
> > +file://0003-gweb-fix-segfault-with-musl-v1.1.21.patch \
> > +"
>
> This fix is not musl specific.
>
> Cheers,
> Andre'
>
> >
> > SRC_URI[md5sum] = "75012084f14fb63a84b116e66c6e94fb"
> > SRC_URI[sha256sum] =
> > "6ce29b3eb0bb16a7387bc609c39455fd13064bdcde5a4d185fab3a0c71946e16"
> > --
> > 2.19.1
> >
>
> --
> ___
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] connman: fix segfault with musl >2.21
On Thu, May 23, 2019 at 02:56:11PM +0100, André Draszik wrote: >... > Also, in connman's gweb.c 'addr' is initialised to NULL. Again, > NULL isn't anything 'returned by getaddrinfo()', so even in glibc > it only works by pure luck. >... Not pure luck, free(NULL) is valid but passing random garbage to freeaddrinfo() would always fail.[1] The critical point is that freeaddrinfo(NULL) working with a C library is only an implementation detail of this specific version of the library, not something a C library has to support according to POSIX. > Cheers, > Andre' cu Adrian [1] unless the garbage happens to be NULL -- "Is there not promise of rain?" Ling Tan asked suddenly out of the darkness. There had been need of rain for many days. "Only a promise," Lao Er said. Pearl S. Buck - Dragon Seed -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH 1/2] piglit: upgrade to latest revision
Can you file a bug to get devtool to state that in the commit message?
Ross
On Thu, 23 May 2019 at 10:27, Alexander Kanavin wrote:
>
> Yes. If devtool drops them, then they are already in the code.
>
> Alex
>
> On Thu, 23 May 2019 at 03:30, Khem Raj wrote:
> >
> > On Wed, May 22, 2019 at 4:02 AM Alexander Kanavin
> > wrote:
> > >
> > > Signed-off-by: Alexander Kanavin
> > > ---
> > > ...UPWARD-only-if-its-defined-in-fenv.h.patch | 75 ---
> > > .../piglit/piglit/format-fix.patch| 69 -
> >
> > are these patches upstreamed ?
> >
> > > meta/recipes-graphics/piglit/piglit_git.bb| 5 +-
> > > 3 files changed, 1 insertion(+), 148 deletions(-)
> > > delete mode 100644
> > > meta/recipes-graphics/piglit/piglit/0001-tests-Use-FE_UPWARD-only-if-its-defined-in-fenv.h.patch
> > > delete mode 100644 meta/recipes-graphics/piglit/piglit/format-fix.patch
> > >
> > > diff --git
> > > a/meta/recipes-graphics/piglit/piglit/0001-tests-Use-FE_UPWARD-only-if-its-defined-in-fenv.h.patch
> > >
> > > b/meta/recipes-graphics/piglit/piglit/0001-tests-Use-FE_UPWARD-only-if-its-defined-in-fenv.h.patch
> > > deleted file mode 100644
> > > index 57eda2e207b..000
> > > ---
> > > a/meta/recipes-graphics/piglit/piglit/0001-tests-Use-FE_UPWARD-only-if-its-defined-in-fenv.h.patch
> > > +++ /dev/null
> > > @@ -1,75 +0,0 @@
> > > -Upstream-Status: Submitted [mailing list]
> > > -Signed-off-by: Ross Burton
> > > -
> > > -From 0e0a2a69261031d55d52b6045990e8982ea12912 Mon Sep 17 00:00:00 2001
> > > -From: Khem Raj
> > > -Date: Sat, 9 Jul 2016 07:52:19 +
> > > -Subject: [PATCH] tests: only run rounding tests if FE_UPWARD is present
> > > -
> > > -On ARM, musl does not define FE_* when the architecture does not have
> > > VFP (which
> > > -is the right interpretation).
> > > -
> > > -As these tests depend on calling fesetround(), skip the test if
> > > FE_UPWARD isn't
> > > -available.
> > > -
> > > -Signed-off-by: Ross Burton
> > >
> > > - tests/general/roundmode-getintegerv.c | 12
> > > - tests/general/roundmode-pixelstore.c | 12
> > > - 2 files changed, 16 insertions(+), 8 deletions(-)
> > > -
> > > -diff --git a/tests/general/roundmode-getintegerv.c
> > > b/tests/general/roundmode-getintegerv.c
> > > -index 28ecfaf55..aa99044a1 100644
> > > a/tests/general/roundmode-getintegerv.c
> > > -+++ b/tests/general/roundmode-getintegerv.c
> > > -@@ -79,13 +79,17 @@ test(float val, int expect)
> > > - void
> > > - piglit_init(int argc, char **argv)
> > > - {
> > > -- int ret;
> > > - bool pass = true;
> > > -- ret = fesetround(FE_UPWARD);
> > > -- if (ret != 0) {
> > > -- printf("Couldn't set rounding mode\n");
> > > -+
> > > -+#ifdef FE_UPWARD
> > > -+ if (fesetround(FE_UPWARD) != 0) {
> > > -+ printf("Setting rounding mode failed\n");
> > > - piglit_report_result(PIGLIT_SKIP);
> > > - }
> > > -+#else
> > > -+ printf("Cannot set rounding mode\n");
> > > -+ piglit_report_result(PIGLIT_SKIP);
> > > -+#endif
> > > -
> > > - pass = test(2.2, 2) && pass;
> > > - pass = test(2.8, 3) && pass;
> > > -diff --git a/tests/general/roundmode-pixelstore.c
> > > b/tests/general/roundmode-pixelstore.c
> > > -index 8a029b257..57ec11c09 100644
> > > a/tests/general/roundmode-pixelstore.c
> > > -+++ b/tests/general/roundmode-pixelstore.c
> > > -@@ -79,13 +79,17 @@ test(float val, int expect)
> > > - void
> > > - piglit_init(int argc, char **argv)
> > > - {
> > > -- int ret;
> > > - bool pass = true;
> > > -- ret = fesetround(FE_UPWARD);
> > > -- if (ret != 0) {
> > > -- printf("Couldn't set rounding mode\n");
> > > -+
> > > -+#ifdef FE_UPWARD
> > > -+ if (fesetround(FE_UPWARD) != 0) {
> > > -+ printf("Setting rounding mode failed\n");
> > > - piglit_report_result(PIGLIT_SKIP);
> > > - }
> > > -+#else
> > > -+ printf("Cannot set rounding mode\n");
> > > -+ piglit_report_result(PIGLIT_SKIP);
> > > -+#endif
> > > -
> > > - pass = test(2.2, 2) && pass;
> > > - pass = test(2.8, 3) && pass;
> > > ---
> > > -2.11.0
> > > -
> > > diff --git a/meta/recipes-graphics/piglit/piglit/format-fix.patch
> > > b/meta/recipes-graphics/piglit/piglit/format-fix.patch
> > > deleted file mode 100644
> > > index 73d539fef27..000
> > > --- a/meta/recipes-graphics/piglit/piglit/format-fix.patch
> > > +++ /dev/null
> > > @@ -1,69 +0,0 @@
> > > -Upstream-Status: Submitted [mailing list]
> > > -Signed-off-by: Ross Burton
> > > -
> > > -From f0c6981322807e179e39ce67aeebd42cf7a54d36 Mon Sep 17 00:00:00 2001
> > > -From: Ross Burton
> > > -Date: Wed, 21 Nov 2018 12:44:36 +
> > > -Subject: [PATCH] arb_texture_view: fix security format warnings
> > > -
> > > -If built with -Werror=format-security then Piglit fails to build:
> > > -
> > >
Re: [OE-core] [PATCH] qemu: IRQ Stats not shown with virsh command
On Thu, 23 May 2019 at 07:48, Jiping Ma wrote: > Issue: LIN10-5872 > > Add IRQ statistic code in kvm/i8259.c. > > (LOCAL REV: NOT UPSTREAM) -- oe-core already included it. Can you remove all these Wind River-specific lines and replace them with a better commit message? Ross -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] package_manager: add help packages based on languages selected in IMAGE_LINGUAS
On 5/23/19 2:39 PM, Burton, Ross wrote: > Why -help- and not -doc? I was thinking it may make better sense to have a variable that DEFAULTS to "*-locale-%s", but can be extended to have other locale based files.. Such as a *-doc-%s, or *-man-%s, etc... (We will -never- get all of the possible ones, unless they all have 'locale' in the names or similar.) --Mark > Ross > > On Thu, 23 May 2019 at 13:30, Andreas Müller wrote: >> >> This is the one-line mentioned in [1]. >> >> [1] >> http://lists.openembedded.org/pipermail/openembedded-devel/2019-May/199769.html >> >> Signed-off-by: Andreas Müller >> --- >> meta/lib/oe/package_manager.py | 1 + >> 1 file changed, 1 insertion(+) >> >> diff --git a/meta/lib/oe/package_manager.py b/meta/lib/oe/package_manager.py >> index 2835c1ddf8..bd2432f6d3 100644 >> --- a/meta/lib/oe/package_manager.py >> +++ b/meta/lib/oe/package_manager.py >> @@ -566,6 +566,7 @@ class PackageManager(object, metaclass=ABCMeta): >> >> for lang in split_linguas: >> globs += " *-locale-%s" % lang >> +globs += " *-help-%s" % lang >> >> if globs is None: >> return >> -- >> 2.20.1 >> >> -- >> ___ >> Openembedded-core mailing list >> Openembedded-core@lists.openembedded.org >> http://lists.openembedded.org/mailman/listinfo/openembedded-core -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH 3/3] elfutils: Fix eu-* utils builds for musl
On Thu, May 23, 2019 at 06:52:23AM -0700, Khem Raj wrote: > On Wed, May 22, 2019 at 11:10 PM Adrian Bunk wrote: > > > On Wed, May 22, 2019 at 04:29:33PM -0700, Khem Raj wrote: > > > On Tue, May 21, 2019 at 11:49 PM Adrian Bunk wrote: > > > > > > > > On Tue, May 21, 2019 at 06:32:28PM -0700, Khem Raj wrote: > > > > >... > > > > > PN and PN-binutils is not empty anymore on musl > > > > >... > > > > > ++#if !defined(FNM_EXTMATCH) > > > > > ++# define FNM_EXTMATCH (1 << 5) /* Use ksh-like extended matching. > > */ > > > > > ++#endif > > > > >... > > > > > > > > This might fix the compilation, but it is horribly wrong. > > > > > > > > It passes a flag to a function that this function does not understand. > > > > > > > > In the best case the function returns with an error on unknown flags. > > > > > > > > In the worst case the function does silently something different from > > > > what the caller expects it to do. > > > > > > > > > > I forgot that the API is provided by libc here so mimic'ing the define > > > infact could > > > cause pain. I think its better to nullify it in this case where it is > > > being used in these > > > files. > > > > Just looking at the usage in src/elfcompress.c, my first impression is > > that this tool is mostly useless without a properly working FNM_EXTMATCH. > > > > Are the eu-* utils actually important in the cases where space is so > > limited that people are forced to use musl instead of glibc? > > > > > They are primarily needed for testing and now we have good auto testing > infra it would be good to build them Known-broken code is not a good basis for testing. When I build elfutils locally on my x86_64 desktop and run "make check" all tests pass (with 3 skipped tests). Adding #undef FNM_EXTMATCH #define FNM_EXTMATCH 0 to the defines in src/elfcompress.c gives me one test failure. That could have easily been a day wasted by someone debugging why this test failed with musl. If these tests are really important the fnmatch replacement from gnulib might help, similar to the other cases where missing functionality is added from gnulib to add functionality for musl builds (e.g. argp, fts). cu Adrian -- "Is there not promise of rain?" Ling Tan asked suddenly out of the darkness. There had been need of rain for many days. "Only a promise," Lao Er said. Pearl S. Buck - Dragon Seed -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] bluez5: remove udev dependency
On 5/23/19 11:12 AM, David Frey wrote:
On 5/22/2019 3:03 PM, Adrian Bunk wrote:
On Wed, May 22, 2019 at 02:55:41PM -0700, David Frey wrote:
udev is an optional dependency of bluez5, so use PACKAGECONFIG to allow
users to decide if they want udev support.
Signed-off-by: David Frey
---
meta/recipes-connectivity/bluez5/bluez5.inc | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/meta/recipes-connectivity/bluez5/bluez5.inc
b/meta/recipes-connectivity/bluez5/bluez5.inc
index aaf2af975d..568de86d6d 100644
--- a/meta/recipes-connectivity/bluez5/bluez5.inc
+++ b/meta/recipes-connectivity/bluez5/bluez5.inc
@@ -6,7 +6,7 @@ LICENSE = "GPLv2+ & LGPLv2.1+"
LIC_FILES_CHKSUM = "file://COPYING;md5=12f884d2ae1ff87c09e5b7ccc2c4ca7e \
file://COPYING.LIB;md5=fb504b67c50331fc78734fed90fb0e09 \
file://src/main.c;beginline=1;endline=24;md5=9bc54b93cd7e17bf03f52513f39f926e"
-DEPENDS = "udev dbus-glib glib-2.0"
+DEPENDS = "dbus-glib glib-2.0"
PROVIDES += "bluez-hcidump"
RPROVIDES_${PN} += "bluez-hcidump"
@@ -43,6 +43,7 @@ PACKAGECONFIG[threads] = "--enable-threads,--disable-threads"
PACKAGECONFIG[deprecated] = "--enable-deprecated,--disable-deprecated"
PACKAGECONFIG[mesh] = "--enable-mesh,--disable-mesh, json-c ell"
PACKAGECONFIG[btpclient] = "--enable-btpclient,--disable-btpclient, ell"
+PACKAGECONFIG[udev] = "--enable-udev,--disable-udev,udev"
...
Unless there is a good reason to do otherwise, this should be added to
the default PACKAGECONFIG so that the set of features enabled by default
stays the same.
I think bluez5 should be built with udev support if udev is going to be
part of the system, but I don't think udev should be brought in as a
dependency of bluez5 by default. How can I express this? I *think*
this is what is done for systemd in the PACKAGECONFIG with this fragment:
${@bb.utils.filter('DISTRO_FEATURES', 'systemd', d)}
I think it already is a dependency by default (before your patch). We
would like this to remain the default for current consumers of bluez
after your patch. I think the easiest way to do this is to have your
patch add "udev" to the default PACKAGECONFIG for the recipe. Then, in
your local.conf/distro.conf/whatever you can override the PACKAGECONFIG
for the recipe to remove udev for your use case.
This way, your patch allow you to have bluez without udev, but you won't
break everyone else by suddenly removing it.
This uses DISTRO_FEATURES and based on this
(https://www.yoctoproject.org/docs/2.7/mega-manual/mega-manual.html#ref-features-distro),
it seems that there is no "udev" distro feature, so I'm not sure how to
proceed.
Thanks,
David
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] Revert "musl: Add TEMP_FAILURE_RETRY from glibc"
On Thu, May 23, 2019 at 07:58:19AM -0700, Khem Raj wrote: >... > Secondly, musl and glibc are under different licenses and importing > code from one to other could taint it. >... How many LOC are considered the legal limit in OE for that? Many of the musl support patches (e.g. [1]) contain more LOC from glibc headers,[2] so if these are considered a problem there might be more legal review ahead. cu Adrian [1] meta/recipes-extended/pam/libpam/0001-Add-support-for-defining-missing-funcitonality.patch [2] https://sourceware.org/git/?p=glibc.git;a=blame;f=string/string.h;hb=0a54e4010fe0085cd36deaff9442a7e88de3270d#l131 -- "Is there not promise of rain?" Ling Tan asked suddenly out of the darkness. There had been need of rain for many days. "Only a promise," Lao Er said. Pearl S. Buck - Dragon Seed -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] bluez5: remove udev dependency
On 5/22/2019 3:03 PM, Adrian Bunk wrote:
> On Wed, May 22, 2019 at 02:55:41PM -0700, David Frey wrote:
>> udev is an optional dependency of bluez5, so use PACKAGECONFIG to allow
>> users to decide if they want udev support.
>>
>> Signed-off-by: David Frey
>> ---
>> meta/recipes-connectivity/bluez5/bluez5.inc | 3 ++-
>> 1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/meta/recipes-connectivity/bluez5/bluez5.inc
>> b/meta/recipes-connectivity/bluez5/bluez5.inc
>> index aaf2af975d..568de86d6d 100644
>> --- a/meta/recipes-connectivity/bluez5/bluez5.inc
>> +++ b/meta/recipes-connectivity/bluez5/bluez5.inc
>> @@ -6,7 +6,7 @@ LICENSE = "GPLv2+ & LGPLv2.1+"
>> LIC_FILES_CHKSUM = "file://COPYING;md5=12f884d2ae1ff87c09e5b7ccc2c4ca7e \
>> file://COPYING.LIB;md5=fb504b67c50331fc78734fed90fb0e09
>> \
>>
>> file://src/main.c;beginline=1;endline=24;md5=9bc54b93cd7e17bf03f52513f39f926e"
>> -DEPENDS = "udev dbus-glib glib-2.0"
>> +DEPENDS = "dbus-glib glib-2.0"
>> PROVIDES += "bluez-hcidump"
>> RPROVIDES_${PN} += "bluez-hcidump"
>>
>> @@ -43,6 +43,7 @@ PACKAGECONFIG[threads] =
>> "--enable-threads,--disable-threads"
>> PACKAGECONFIG[deprecated] = "--enable-deprecated,--disable-deprecated"
>> PACKAGECONFIG[mesh] = "--enable-mesh,--disable-mesh, json-c ell"
>> PACKAGECONFIG[btpclient] = "--enable-btpclient,--disable-btpclient, ell"
>> +PACKAGECONFIG[udev] = "--enable-udev,--disable-udev,udev"
>> ...
>
> Unless there is a good reason to do otherwise, this should be added to
> the default PACKAGECONFIG so that the set of features enabled by default
> stays the same.
I think bluez5 should be built with udev support if udev is going to be
part of the system, but I don't think udev should be brought in as a
dependency of bluez5 by default. How can I express this? I *think*
this is what is done for systemd in the PACKAGECONFIG with this fragment:
${@bb.utils.filter('DISTRO_FEATURES', 'systemd', d)}
This uses DISTRO_FEATURES and based on this
(https://www.yoctoproject.org/docs/2.7/mega-manual/mega-manual.html#ref-features-distro),
it seems that there is no "udev" distro feature, so I'm not sure how to
proceed.
Thanks,
David
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] bluez5: remove udev dependency
On 5/23/2019 9:30 AM, Joshua Watt wrote:
>
> On 5/23/19 11:12 AM, David Frey wrote:
>> On 5/22/2019 3:03 PM, Adrian Bunk wrote:
>>> On Wed, May 22, 2019 at 02:55:41PM -0700, David Frey wrote:
udev is an optional dependency of bluez5, so use PACKAGECONFIG to allow
users to decide if they want udev support.
Signed-off-by: David Frey
---
meta/recipes-connectivity/bluez5/bluez5.inc | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/meta/recipes-connectivity/bluez5/bluez5.inc
b/meta/recipes-connectivity/bluez5/bluez5.inc
index aaf2af975d..568de86d6d 100644
--- a/meta/recipes-connectivity/bluez5/bluez5.inc
+++ b/meta/recipes-connectivity/bluez5/bluez5.inc
@@ -6,7 +6,7 @@ LICENSE = "GPLv2+ & LGPLv2.1+"
LIC_FILES_CHKSUM =
"file://COPYING;md5=12f884d2ae1ff87c09e5b7ccc2c4ca7e \
file://COPYING.LIB;md5=fb504b67c50331fc78734fed90fb0e09 \
file://src/main.c;beginline=1;endline=24;md5=9bc54b93cd7e17bf03f52513f39f926e"
-DEPENDS = "udev dbus-glib glib-2.0"
+DEPENDS = "dbus-glib glib-2.0"
PROVIDES += "bluez-hcidump"
RPROVIDES_${PN} += "bluez-hcidump"
@@ -43,6 +43,7 @@ PACKAGECONFIG[threads] =
"--enable-threads,--disable-threads"
PACKAGECONFIG[deprecated] =
"--enable-deprecated,--disable-deprecated"
PACKAGECONFIG[mesh] = "--enable-mesh,--disable-mesh, json-c ell"
PACKAGECONFIG[btpclient] =
"--enable-btpclient,--disable-btpclient, ell"
+PACKAGECONFIG[udev] = "--enable-udev,--disable-udev,udev"
...
>>> Unless there is a good reason to do otherwise, this should be added to
>>> the default PACKAGECONFIG so that the set of features enabled by default
>>> stays the same.
>> I think bluez5 should be built with udev support if udev is going to be
>> part of the system, but I don't think udev should be brought in as a
>> dependency of bluez5 by default. How can I express this? I *think*
>> this is what is done for systemd in the PACKAGECONFIG with this fragment:
>> ${@bb.utils.filter('DISTRO_FEATURES', 'systemd', d)}
>
> I think it already is a dependency by default (before your patch). We
> would like this to remain the default for current consumers of bluez
> after your patch. I think the easiest way to do this is to have your
> patch add "udev" to the default PACKAGECONFIG for the recipe. Then, in
> your local.conf/distro.conf/whatever you can override the PACKAGECONFIG
> for the recipe to remove udev for your use case.
>
> This way, your patch allow you to have bluez without udev, but you won't
> break everyone else by suddenly removing it.
I have submitted v2 based on your feedback.
Thanks,
David
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH v2] bluez5: remove udev dependency
udev is an optional dependency of bluez5, so use PACKAGECONFIG to allow
users to decide if they want udev support.
Signed-off-by: David Frey
---
meta/recipes-connectivity/bluez5/bluez5.inc | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/meta/recipes-connectivity/bluez5/bluez5.inc
b/meta/recipes-connectivity/bluez5/bluez5.inc
index aaf2af975d..93d1b4d8b0 100644
--- a/meta/recipes-connectivity/bluez5/bluez5.inc
+++ b/meta/recipes-connectivity/bluez5/bluez5.inc
@@ -6,7 +6,7 @@ LICENSE = "GPLv2+ & LGPLv2.1+"
LIC_FILES_CHKSUM = "file://COPYING;md5=12f884d2ae1ff87c09e5b7ccc2c4ca7e \
file://COPYING.LIB;md5=fb504b67c50331fc78734fed90fb0e09 \
file://src/main.c;beginline=1;endline=24;md5=9bc54b93cd7e17bf03f52513f39f926e"
-DEPENDS = "udev dbus-glib glib-2.0"
+DEPENDS = "dbus-glib glib-2.0"
PROVIDES += "bluez-hcidump"
RPROVIDES_${PN} += "bluez-hcidump"
@@ -22,6 +22,7 @@ PACKAGECONFIG ??= "obex-profiles \
hog-profiles \
tools \
deprecated \
+udev \
"
PACKAGECONFIG[obex-profiles] = "--enable-obex,--disable-obex,libical"
PACKAGECONFIG[readline] = "--enable-client,--disable-client,readline,"
@@ -43,6 +44,7 @@ PACKAGECONFIG[threads] = "--enable-threads,--disable-threads"
PACKAGECONFIG[deprecated] = "--enable-deprecated,--disable-deprecated"
PACKAGECONFIG[mesh] = "--enable-mesh,--disable-mesh, json-c ell"
PACKAGECONFIG[btpclient] = "--enable-btpclient,--disable-btpclient, ell"
+PACKAGECONFIG[udev] = "--enable-udev,--disable-udev,udev"
SRC_URI = "\
${KERNELORG_MIRROR}/linux/bluetooth/bluez-${PV}.tar.xz \
--
2.21.0
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [AUH] Upgrade status: 2019-05-23
Recipe upgrade statistics: * Failed (devtool error): 38 apt, 1.8.1, Aníbal Limón openssh, 8.0p1, Armin Kuster gettext, 0.20.1, Robert Yang vulkan, 1.1.106.0, Ross Burton sysklogd, 1.6, Chen Qi ifupdown, 0.8.35, Anuj Mittal sysvinit, 2.94, Ross Burton libidn2, 2.1.1a, Ross Burton flex, 2.6.4, Chen Qi gdk-pixbuf, 2.38.1, Ross Burton ruby, 2.6.3, Ross Burton iputils, s20190515, Changhyeok Bae ltp, 20190517, Yi Zhao e2fsprogs, 1.45.1, Robert Yang strace, 5.0, Robert Yang go-dep, 0.5.3, Otavio Salvador vulkan-demos, git-new-commits-available, Ross Burton sed, 4.7, Chen Qi puzzles, 0.0-new-commits-available, Anuj Mittal texinfo, 6.6, Anuj Mittal kmscube, git-new-commits-available, Carlos Rafael Giani liburcu, 0.11.0, Alexander Kanavin attr, 2.4.48, Chen Qi rng-tools, 6.7, Anuj Mittal glib-2.0, 2.60.3, Anuj Mittal acl, 2.2.53, Chen Qi util-linux, 2.33.2, Chen Qi squashfs-tools, 4.3-new-commits-available, Robert Yang nfs-utils, 2.3.4, Robert Yang lsbinitscripts, 10.01, Ross Burton gstreamer1.0-plugins-bad, 1.16.0, Anuj Mittal quilt, 0.66, Robert Yang dpkg, 1.19.6, Aníbal Limón gstreamer1.0-plugins-base, 1.16.0, Anuj Mittal build-compare, 2015.02.10-new-commits-available, Paul Eggleton libical, 3.0.5, Ross Burton gtk+3, 3.24.8, Ross Burton gstreamer1.0-rtsp-server, 1.16.0, Anuj Mittal * Succeeded: 14 iproute2, 5.1.0, Changhyeok Bae dbus-test, 1.12.14, Chen Qi piglit, 1.0-new-commits-available, Ross Burton curl, 7.65.0, Armin Kuster dnf, 4.2.6, Alexander Kanavin dbus, 1.12.14, Chen Qi libva, 2.4.1, Anuj Mittal file, 5.37, Yi Zhao vim-tiny, 8.1.1364, Tom Rini vte, 0.56.3, Anuj Mittal bison, 3.4, Chen Qi vim, 8.1.1364, Tom Rini eudev, 3.2.8, Anuj Mittal libdnf, 0.33.0, Alexander Kanavin * Failed(other errors): 1 gcc-source-8.3.0, 9.1.0, Khem Raj * Failed(do_compile): 19 mesa, 19.0.5, Otavio Salvador gstreamer1.0-vaapi, 1.16.0, Anuj Mittal gsettings-desktop-schemas, 3.32.0, Anuj Mittal gstreamer1.0-python, 1.16.0, Anuj Mittal mkfontscale, 1.2.1, Armin Kuster autoconf-archive, 2019.01.06, Robert Yang cmake, 3.14.4, Pascal Bach gst-validate, 1.16.0, Anuj Mittal mtd-utils, 2.1.0, Denys Dmytriyenko prelink, 1.0-new-commits-available, Mark Hatle ethtool, 5.1, Changhyeok Bae gtk-doc, 1.30, Alexander Kanavin gnutls, 3.6.7.1, Armin Kuster gstreamer1.0-libav, 1.16.0, Anuj Mittal gstreamer1.0-plugins-ugly, 1.16.0, Anuj Mittal boost, 1.70.0, Alexander Kanavin gstreamer1.0, 1.16.0, Anuj Mittal gstreamer1.0-plugins-good, 1.16.0, Anuj Mittal gstreamer1.0-omx, 1.16.0, Anuj Mittal TOTAL: attempted=72 succeeded=14(19.44%) failed=58(80.56%) Recipe upgrade statistics per Maintainer: Changhyeok Bae -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] Revert "musl: Add TEMP_FAILURE_RETRY from glibc"
On Thu, May 23, 2019 at 9:06 AM Adrian Bunk wrote: > > On Thu, May 23, 2019 at 07:58:19AM -0700, Khem Raj wrote: > >... > > Secondly, musl and glibc are under different licenses and importing > > code from one to other could taint it. > >... > > How many LOC are considered the legal limit in OE for that? > Just make it easy where it can. OE does not guarantee any of that. > Many of the musl support patches (e.g. [1]) contain more LOC from glibc > headers,[2] so if these are considered a problem there might be more > legal review ahead. > Yes I am aware. > cu > Adrian > > [1] > meta/recipes-extended/pam/libpam/0001-Add-support-for-defining-missing-funcitonality.patch > [2] > https://sourceware.org/git/?p=glibc.git;a=blame;f=string/string.h;hb=0a54e4010fe0085cd36deaff9442a7e88de3270d#l131 > > -- > >"Is there not promise of rain?" Ling Tan asked suddenly out > of the darkness. There had been need of rain for many days. >"Only a promise," Lao Er said. >Pearl S. Buck - Dragon Seed > -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH 1/2] texinfo-dummy-native: A little clean up of template.py
This is mainly whitespace clean up, plus using the with statement when
writing files.
Signed-off-by: Peter Kjellerstedt
---
.../texinfo-dummy/template.py | 63 +--
1 file changed, 30 insertions(+), 33 deletions(-)
diff --git
a/meta/recipes-extended/texinfo-dummy-native/texinfo-dummy/template.py
b/meta/recipes-extended/texinfo-dummy-native/texinfo-dummy/template.py
index e369f74455..fcc28548af 100644
--- a/meta/recipes-extended/texinfo-dummy-native/texinfo-dummy/template.py
+++ b/meta/recipes-extended/texinfo-dummy-native/texinfo-dummy/template.py
@@ -33,21 +33,20 @@ import sys, os
olong = "--output="
Elong = "--macro-expand="
-
-this_binary = sys.argv[0].split ("/")[-1]
+this_binary = sys.argv[0].split("/")[-1]
# To be outputted if functionality that hasn't been stubbed yet is invoked.
stub_msg = """
-This stand-in version of %s is not yet fully capable of emulating the real
-version from the GNU texinfo suite. If you see this message, file a bug report
-with details on the recipe that failed.
+This stand-in version of %s is not yet fully capable of emulating
+the real version from the GNU texinfo suite. If you see this message, file a
+bug report with details on the recipe that failed.
""" % this_binary
# Autotools setups query the version, so this is actually necessary. Some of
# them (lookin' at you, glibc) actually look for the substring "GNU texinfo,"
# so we put that substring in there without actually telling a lie.
-version_str = """ %s (fake texinfo, emulating GNU texinfo) 5.2
-
+version_str = """%s (fake texinfo, emulating GNU texinfo) 5.2
+
Super amazing version which is totally not fake in any way whatsoever.
Copyright (C) 2014 Intel Corp. Distributed under the terms of the MIT
license.
@@ -55,62 +54,61 @@ license.
simple_binaries = "pod2texi texi2dvi pdftexi2dvi texindex texi2pdf \
txixml2texi install-info ginstall-info \
- update-info-dir".split ()
+ update-info-dir".split()
# These utilities use a slightly different set of options and flags.
-complex_binaries = "makeinfo texi2any".split ()
+complex_binaries = "makeinfo texi2any".split()
valid_binaries = simple_binaries + complex_binaries
# For generating blank output files.
-def touch_file (path):
-f = open (path, "w")
-f.close ()
+def touch_file(path):
+with open(path, "w"):
+pass
assert this_binary in valid_binaries, \
- this_binary + " is not one of " + ', '.join (valid_binaries)
+ this_binary + " is not one of " + ', '.join(valid_binaries)
if "--version" in sys.argv:
print(version_str)
-sys.exit (0)
+sys.exit(0)
# For debugging
log_interceptions = False
if log_interceptions:
-f = open ("/tmp/intercepted_" + this_binary, "a")
-f.write (' '.join ([this_binary] + sys.argv[1:]) + '\n')
-f.close ()
+with open("/tmp/intercepted_" + this_binary, "a") as f:
+f.write(' '.join([this_binary] + sys.argv[1:]) + '\n')
# Look through the options and flags, and if necessary, touch any output
# files.
arg_idx = 1
-while arg_idx < len (sys.argv):
+while arg_idx < len(sys.argv):
arg = sys.argv [arg_idx]
-
+
if arg == "--":
break
-
+
# Something like -I . can result in a need for this (specifically the .)
-elif len (arg) < 2:
+elif len(arg) < 2:
pass
-
+
# Check if -o or --output is specified. These can be used at most once.
-elif arg[0] == '-' and arg[1] != '-' and arg[len (arg) - 1] == 'o':
-touch_file (sys.argv[arg_idx + 1])
-sys.exit (0)
-elif arg.startswith (olong):
-touch_file (arg.split ("=")[1])
-sys.exit (0)
-
+elif arg[0] == '-' and arg[1] != '-' and arg[len(arg) - 1] == 'o':
+touch_file(sys.argv[arg_idx + 1])
+sys.exit(0)
+elif arg.startswith(olong):
+touch_file(arg.split("=")[1])
+sys.exit(0)
+
# Check for functionality that isn't implemented yet.
else:
assert arg[0] != '-' or arg[1] == '-' or 'E' not in arg or \
this_binary in simple_binaries, \
"-E option not yet supported" + stub_msg
-
-assert not arg.startswith (Elong), \
+
+assert not arg.startswith(Elong), \
Elong[:-1] + " option not yet supported" + stub_msg
-
+
arg_idx += 1
# The -o/--output option overrides the default. For makeinfo and texi2any,
@@ -119,4 +117,3 @@ while arg_idx < len (sys.argv):
assert this_binary in simple_binaries, \
"Don't know how to get default output file name from input file!" + \
stub_msg
-
--
2.21.0
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH 2/2] texinfo-dummy-native: Rewrite template.py to use argparse
The original version of template.py parses the arguments manually. This
fails when looking for the -E option if, e.g., an -I option is specified
without any space before its argument, and that argument contains the
letter 'E'.
A minor difference to the original version is that it parsed the
arguments in the order they were specified on the command line whereas
this version will always handle -E before -o.
Signed-off-by: Peter Kjellerstedt
---
.../texinfo-dummy/template.py | 55 ++-
1 file changed, 18 insertions(+), 37 deletions(-)
diff --git
a/meta/recipes-extended/texinfo-dummy-native/texinfo-dummy/template.py
b/meta/recipes-extended/texinfo-dummy-native/texinfo-dummy/template.py
index fcc28548af..86c7c1811a 100644
--- a/meta/recipes-extended/texinfo-dummy-native/texinfo-dummy/template.py
+++ b/meta/recipes-extended/texinfo-dummy-native/texinfo-dummy/template.py
@@ -28,10 +28,8 @@
# of the executable from argv[0] and emulate the corresponding program, so
# multiple copies of this script will exist under different names.
-import sys, os
+import sys, os, argparse
-olong = "--output="
-Elong = "--macro-expand="
this_binary = sys.argv[0].split("/")[-1]
@@ -61,18 +59,9 @@ complex_binaries = "makeinfo texi2any".split()
valid_binaries = simple_binaries + complex_binaries
-# For generating blank output files.
-def touch_file(path):
-with open(path, "w"):
-pass
-
assert this_binary in valid_binaries, \
this_binary + " is not one of " + ', '.join(valid_binaries)
-if "--version" in sys.argv:
-print(version_str)
-sys.exit(0)
-
# For debugging
log_interceptions = False
if log_interceptions:
@@ -81,35 +70,27 @@ if log_interceptions:
# Look through the options and flags, and if necessary, touch any output
# files.
-arg_idx = 1
-while arg_idx < len(sys.argv):
-arg = sys.argv [arg_idx]
-
-if arg == "--":
-break
+p = argparse.ArgumentParser()
+if this_binary in complex_binaries:
+p.add_argument('-E', '--macro-expand', metavar='FILE')
+p.add_argument('-o', '--output', metavar='DEST')
+p.add_argument('--version', action='store_true')
-# Something like -I . can result in a need for this (specifically the .)
-elif len(arg) < 2:
-pass
-
-# Check if -o or --output is specified. These can be used at most once.
-elif arg[0] == '-' and arg[1] != '-' and arg[len(arg) - 1] == 'o':
-touch_file(sys.argv[arg_idx + 1])
-sys.exit(0)
-elif arg.startswith(olong):
-touch_file(arg.split("=")[1])
-sys.exit(0)
+args, unknown = p.parse_known_args()
-# Check for functionality that isn't implemented yet.
-else:
-assert arg[0] != '-' or arg[1] == '-' or 'E' not in arg or \
- this_binary in simple_binaries, \
- "-E option not yet supported" + stub_msg
+if args.version:
+print(version_str)
+sys.exit(0)
-assert not arg.startswith(Elong), \
- Elong[:-1] + " option not yet supported" + stub_msg
+# Check for functionality that isn't implemented yet.
+assert not getattr(args, 'macro_expand', None), \
+"-E/--macro-expand option not yet supported" + stub_msg
-arg_idx += 1
+# Check if -o or --output is specified.
+if args.output:
+with open(args.output, 'w'):
+pass
+sys.exit(0)
# The -o/--output option overrides the default. For makeinfo and texi2any,
# that default is to look for a @setfilename command in the input file.
--
2.21.0
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH v2] connman: fix segfault with musl >v1.1.21
On 5/23/19 6:26 AM, Nicola Lunghi wrote:
musl > v1.1.21 changed the implementation of the freeaddrinfo() function
not allowing anymore to pass null pointers to it.
This was causing
Looks ok. We had similar issue with nfs-utils
https://patchwork.openembedded.org/patch/156772/
a segmentation fault in connman.
Signed-off-by: Nicola Lunghi
---
...-gweb-fix-segfault-with-musl-v1.1.21.patch | 34 +++
.../connman/connman_1.37.bb | 1 +
2 files changed, 35 insertions(+)
create mode 100644
meta/recipes-connectivity/connman/connman/0001-gweb-fix-segfault-with-musl-v1.1.21.patch
diff --git
a/meta/recipes-connectivity/connman/connman/0001-gweb-fix-segfault-with-musl-v1.1.21.patch
b/meta/recipes-connectivity/connman/connman/0001-gweb-fix-segfault-with-musl-v1.1.21.patch
new file mode 100644
index 00..43b43bc9f8
--- /dev/null
+++
b/meta/recipes-connectivity/connman/connman/0001-gweb-fix-segfault-with-musl-v1.1.21.patch
@@ -0,0 +1,34 @@
+From f0a8c69971b30ea7ca255bb885fdd1179fa5d298 Mon Sep 17 00:00:00 2001
+From: Nicola Lunghi
+Date: Thu, 23 May 2019 07:55:25 +0100
+Subject: [PATCH] gweb: fix segfault with musl v1.1.21
+
+In musl > v1.1.21 freeaddrinfo() implementation changed and
+was causing a segmentation fault on recent Yocto using musl.
+
+See this commit:
+
+
https://git.musl-libc.org/cgit/musl/commit/src/network/freeaddrinfo.c?id=d1395c43c019aec6b855cf3c656bf47c8a719e7f
+
+Upstream-Status: Submitted
+---
+ gweb/gweb.c | 3 ++-
+ 1 file changed, 2 insertions(+), 1 deletion(-)
+
+diff --git a/gweb/gweb.c b/gweb/gweb.c
+index 393afe0a..12fcb1d8 100644
+--- a/gweb/gweb.c
b/gweb/gweb.c
+@@ -1274,7 +1274,8 @@ static bool is_ip_address(const char *host)
+ addr = NULL;
+
+ result = getaddrinfo(host, NULL, , );
+- freeaddrinfo(addr);
++ if(!result)
++ freeaddrinfo(addr);
+
+ return result == 0;
+ }
+--
+2.19.1
+
diff --git a/meta/recipes-connectivity/connman/connman_1.37.bb
b/meta/recipes-connectivity/connman/connman_1.37.bb
index 2cf904cd85..12650329c5 100644
--- a/meta/recipes-connectivity/connman/connman_1.37.bb
+++ b/meta/recipes-connectivity/connman/connman_1.37.bb
@@ -3,6 +3,7 @@ require connman.inc
SRC_URI = "${KERNELORG_MIRROR}/linux/network/${BPN}/${BP}.tar.xz \
file://0001-plugin.h-Change-visibility-to-default-for-debug-symb.patch \
file://0001-connman.service-stop-systemd-resolved-when-we-use-co.patch \
+file://0001-gweb-fix-segfault-with-musl-v1.1.21.patch \
file://connman \
file://no-version-scripts.patch \
"
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH v3 1/2] local.conf.sample: make systemd as default init manager
> -Original Message- > From: openembedded-core-boun...@lists.openembedded.org core-boun...@lists.openembedded.org> On Behalf Of > kai.k...@windriver.com > Sent: den 23 maj 2019 10:26 > To: richard.pur...@linuxfoundation.org > Cc: openembedded-core@lists.openembedded.org > Subject: [OE-core] [PATCH v3 1/2] local.conf.sample: make systemd as > default init manager > > From: Kai Kang > > Move configurations from local.conf.sample.extended to local.conf.sample > to make systemd as default init manager for poky. If we're going to change the default init manager to be systemd, wouldn't it be more appropriate to change the real default values in bitbake.conf and packagegroup-core-boot.bb? And then include an example in local.conf.sample.extended to show how to configure sysvinit as init manager? > > [YOCTO #13031] > > Signed-off-by: Kai Kang > --- > meta-poky/conf/local.conf.sample | 9 + > meta-poky/conf/local.conf.sample.extended | 9 - > 2 files changed, 9 insertions(+), 9 deletions(-) > > diff --git a/meta-poky/conf/local.conf.sample b/meta- > poky/conf/local.conf.sample > index 9068e567dc..3a07105e44 100644 > --- a/meta-poky/conf/local.conf.sample > +++ b/meta-poky/conf/local.conf.sample > @@ -249,3 +249,12 @@ PACKAGECONFIG_append_pn-nativesdk-qemu = " sdl" > # track the version of this file when it was generated. This can > safely be ignored if > # this doesn't mean anything to you. > CONF_VERSION = "1" > + > +# > +# Use systemd for system initialization > +# > +DISTRO_FEATURES_append = " systemd" > +DISTRO_FEATURES_BACKFILL_CONSIDERED += "sysvinit" > +VIRTUAL-RUNTIME_login_manager = "shadow-base" > +VIRTUAL-RUNTIME_init_manager = "systemd" > +VIRTUAL-RUNTIME_initscripts = "systemd-compat-units" > diff --git a/meta-poky/conf/local.conf.sample.extended b/meta- > poky/conf/local.conf.sample.extended > index 26603debe6..e20fc5dbf3 100644 > --- a/meta-poky/conf/local.conf.sample.extended > +++ b/meta-poky/conf/local.conf.sample.extended > @@ -377,12 +377,3 @@ DISTRO_FEATURES_remove = "x11" > #VIRTUAL-RUNTIME_initscripts = "initscripts" > #VIRTUAL-RUNTIME_keymaps = "keymaps" > #DISTRO_FEATURES_BACKFILL_CONSIDERED += "sysvinit" > - > -# > -# Use systemd for system initialization > -# > -#DISTRO_FEATURES_append = " systemd" > -#DISTRO_FEATURES_BACKFILL_CONSIDERED += "sysvinit" > -#VIRTUAL-RUNTIME_login_manager = "shadow-base" > -#VIRTUAL-RUNTIME_init_manager = "systemd" > -#VIRTUAL-RUNTIME_initscripts = "systemd-compat-units" > -- > 2.18.0 > > -- > ___ > Openembedded-core mailing list > Openembedded-core@lists.openembedded.org > http://lists.openembedded.org/mailman/listinfo/openembedded-core -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH v3 1/2] local.conf.sample: make systemd as default init manager
On Thu, May 23, 2019 at 1:41 PM Peter Kjellerstedt < peter.kjellerst...@axis.com> wrote: > > -Original Message- > > From: openembedded-core-boun...@lists.openembedded.org > core-boun...@lists.openembedded.org> On Behalf Of > > kai.k...@windriver.com > > Sent: den 23 maj 2019 10:26 > > To: richard.pur...@linuxfoundation.org > > Cc: openembedded-core@lists.openembedded.org > > Subject: [OE-core] [PATCH v3 1/2] local.conf.sample: make systemd as > > default init manager > > > > From: Kai Kang > > > > Move configurations from local.conf.sample.extended to local.conf.sample > > to make systemd as default init manager for poky. > > If we're going to change the default init manager to be systemd, wouldn't > it be more appropriate to change the real default values in bitbake.conf > and packagegroup-core-boot.bb? And then include an example in > local.conf.sample.extended to show how to configure sysvinit as init > manager? That would change it for Oe-core and other distributions as well which is not the intention > > > > > > [YOCTO #13031] > > > > Signed-off-by: Kai Kang > > --- > > meta-poky/conf/local.conf.sample | 9 + > > meta-poky/conf/local.conf.sample.extended | 9 - > > 2 files changed, 9 insertions(+), 9 deletions(-) > > > > diff --git a/meta-poky/conf/local.conf.sample b/meta- > > poky/conf/local.conf.sample > > index 9068e567dc..3a07105e44 100644 > > --- a/meta-poky/conf/local.conf.sample > > +++ b/meta-poky/conf/local.conf.sample > > @@ -249,3 +249,12 @@ PACKAGECONFIG_append_pn-nativesdk-qemu = " sdl" > > # track the version of this file when it was generated. This can > > safely be ignored if > > # this doesn't mean anything to you. > > CONF_VERSION = "1" > > + > > +# > > +# Use systemd for system initialization > > +# > > +DISTRO_FEATURES_append = " systemd" > > +DISTRO_FEATURES_BACKFILL_CONSIDERED += "sysvinit" > > +VIRTUAL-RUNTIME_login_manager = "shadow-base" > > +VIRTUAL-RUNTIME_init_manager = "systemd" > > +VIRTUAL-RUNTIME_initscripts = "systemd-compat-units" > > diff --git a/meta-poky/conf/local.conf.sample.extended b/meta- > > poky/conf/local.conf.sample.extended > > index 26603debe6..e20fc5dbf3 100644 > > --- a/meta-poky/conf/local.conf.sample.extended > > +++ b/meta-poky/conf/local.conf.sample.extended > > @@ -377,12 +377,3 @@ DISTRO_FEATURES_remove = "x11" > > #VIRTUAL-RUNTIME_initscripts = "initscripts" > > #VIRTUAL-RUNTIME_keymaps = "keymaps" > > #DISTRO_FEATURES_BACKFILL_CONSIDERED += "sysvinit" > > - > > -# > > -# Use systemd for system initialization > > -# > > -#DISTRO_FEATURES_append = " systemd" > > -#DISTRO_FEATURES_BACKFILL_CONSIDERED += "sysvinit" > > -#VIRTUAL-RUNTIME_login_manager = "shadow-base" > > -#VIRTUAL-RUNTIME_init_manager = "systemd" > > -#VIRTUAL-RUNTIME_initscripts = "systemd-compat-units" > > -- > > 2.18.0 > > > > -- > > ___ > > Openembedded-core mailing list > > Openembedded-core@lists.openembedded.org > > http://lists.openembedded.org/mailman/listinfo/openembedded-core > -- > ___ > Openembedded-core mailing list > Openembedded-core@lists.openembedded.org > http://lists.openembedded.org/mailman/listinfo/openembedded-core > -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] Should systemd be marked as incompatible with musl?
On 05/23/2019 08:22 PM, Burton, Ross wrote:
On Thu, 23 May 2019 at 11:34, Adrian Bunk wrote:
systemd is fundamentally Linux-only and not portable to other kernels.
systemd upstream is using glibc extensions not present in other
C libraries.
systemd upstream is accepting technically correct patches that help
building with musl, but there is no interest upstream in keeping systemd
working with non-glibc C libraries.
The way things currently go, systemd/musl will require an ever-growing
amount of not upstreamable patches - and this is not sustainable.
I think I have to agree with you: several of the extensions are for
security purposes, so we're potentially actively introducing issues.
Ross
I suggest we only apply these musl patches when the libc is musl.
SRC_URI += "${@d.getVar('SRC_URI_MUSL') if d.getVar('TCLIBC') == 'musl'
else ''}"
In this way, these patches do not affect glibc users. People who
actually use musl will test these patches and maybe improve them if they
really care about it.
For example, the following patches are written by me.
0017-Do-not-disable-buffering-when-writing-to-oom_score_a.patch and
0001-do-not-disable-buffer-in-writing-files.patch
These two patches are to solve runtime problems in musl based system. I
knew there's something different in buffering mechanism between glibc
and musl that causes this problem, but I really did not want to spend
time digging into it. And as systemd was using unbuffered way for
writing files for a long time, I just disabled them.
As Adrian and Ross pointed out, these patches are introducing potential
security problems. I agree that we should do something about it. At a
minimum, the glibc based system should not be affected.
I'll send out patch and please help review it.
Best Regards,
Chen Qi
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH] avahi: Add PACKAGECONFIG for libdns_sd
Adds a PACKAGECONFIG option to enable the libdns_sd compatibility
library.
Signed-off-by: Joshua Watt
---
meta/recipes-connectivity/avahi/avahi.inc | 7 +++
1 file changed, 7 insertions(+)
diff --git a/meta/recipes-connectivity/avahi/avahi.inc
b/meta/recipes-connectivity/avahi/avahi.inc
index 8339e451f5..94fe6a16b6 100644
--- a/meta/recipes-connectivity/avahi/avahi.inc
+++ b/meta/recipes-connectivity/avahi/avahi.inc
@@ -36,6 +36,7 @@ PACKAGECONFIG ??= "dbus ${AVAHI_GTK}"
PACKAGECONFIG[dbus] = "--enable-dbus,--disable-dbus,dbus"
PACKAGECONFIG[gtk] = "--enable-gtk,--disable-gtk,gtk+"
PACKAGECONFIG[gtk3] = "--enable-gtk3,--disable-gtk3,gtk+3"
+PACKAGECONFIG[libdns_sd] = "--enable-compat-libdns_sd --enable-dbus,,dbus"
inherit autotools pkgconfig gettext gobject-introspection
@@ -77,3 +78,9 @@ do_install() {
test -d ${D}${datadir}/dbus-1 && rmdir --ignore-fail-on-non-empty
${D}${datadir}/dbus-1
rm -rf ${D}${libdir}/avahi
}
+
+PACKAGES =+ "${@bb.utils.contains("PACKAGECONFIG", "libdns_sd",
"libavahi-compat-libdnssd", "", d)}"
+
+FILES_libavahi-compat-libdnssd = "${libdir}/libdns_sd.so.*"
+
+RPROVIDES_libavahi-compat-libdnssd = "libdns-sd"
--
2.21.0
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] possible issue with buildhistory?
On 5/22/19 11:53 PM, Nicolas Dechesne wrote:
hi,
not sure if it's been discussed reported before. but i ended up in a
situation where a package is "empty" (e.g. it's not created) even
though "files-in-packages.txt" in its buildhistory folder shows it
should contain a file.
I have a synthetic example:
$ cat buildhistory/packages/aarch64-linaro-linux/helloworld/latest
PV = 1.0
PR = r0
DEPENDS = autoconf-native automake-native gnu-config-native
libtool-cross libtool-native virtual/aarch64-linaro-linux-compilerlibs
virtual/aarch64-linaro-linux-gcc virtual/libc
PACKAGES = helloworld-src helloworld-dbg helloworld-staticdev
helloworld-dev helloworld-doc helloworld-locale helloworld
goodbyeworld
LAYER = meta-rpb
then:
$ cat
buildhistory/packages/aarch64-linaro-linux/helloworld/goodbyeworld/files-in-package.txt
drwxr-xr-x root root 4096 ./usr
drwxr-xr-x root root 4096 ./usr/bin
-rwxr-xr-x root root 6144 ./usr/bin/goodbyeworld
However goodbyeworld package does not exist, and /usr/bin/goodbyeworld
is in the helloworld package:
$ dpkg -c tmp-rpb-glibc/deploy/ipk/aarch64/helloworld_1.0-r0_aarch64.ipk
drwxrwxrwx root/root 0 2019-05-23 08:25 ./
drwxr-xr-x root/root 0 2019-05-23 08:25 ./usr/
drwxr-xr-x root/root 0 2019-05-23 08:25 ./usr/bin/
-rwxr-xr-x root/root 6144 2019-05-23 08:25 ./usr/bin/goodbyeworld
-rwxr-xr-x root/root 6144 2019-05-23 08:25 ./usr/bin/helloworld
It is worth saying that the file should indeed be in helloworld
package since there was a bug in my recipe! So buildhistory is wrong
here.
The recipe is a simple one:
SRC_URI = "file://helloworld.c \
file://goodbyeworld.c \
file://Makefile.am \
file://configure.ac \
file://COPYING"
S = "${WORKDIR}"
inherit autotools
PACKAGES += "goodbyeworld"
FILES_goodbyeworld += "${bindir}/goodbyeworld"
Clearly the recipe has a bug, and if I do this instead:
PACKAGES =+ "goodbyeworld"
/usr/bin/goodbyeworld ends up in the right package... however I think
in the previous case, buildhistory should have shown
/usb/bin/goodbyeworld in helloworld package, not goodbyeworld...
I was testing on warrior, but i can try to reproduce on master if needed..
I think this could be happening since removing empty package is an
optimization which is not considered by buildhistory. Its worth a bug
report.
thanks!
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH 1/1] systemd: avoid musl specific patches affect glibc systems
On 5/23/19 7:17 PM, Chen Qi wrote:
systemd upstream only care about glibc. We made musl specific
patches so that systemd could work. But currently these patches
contain potential security issues.
So apply these patches only when the libc is musl.
LGTM
Signed-off-by: Chen Qi
---
meta/recipes-core/systemd/systemd_242.bb | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/meta/recipes-core/systemd/systemd_242.bb
b/meta/recipes-core/systemd/systemd_242.bb
index 2dda0d0..adb592f 100644
--- a/meta/recipes-core/systemd/systemd_242.bb
+++ b/meta/recipes-core/systemd/systemd_242.bb
@@ -27,7 +27,7 @@ SRC_URI += "file://touchscreen.rules \
"
# patches needed by musl
-SRC_URI += "${SRC_URI_MUSL}"
+SRC_URI += "${@d.getVar('SRC_URI_MUSL') if d.getVar('TCLIBC') == 'musl' else
''}"
SRC_URI_MUSL =
"file://0001-Use-getenv-when-secure-versions-are-not-available.patch \
file://0002-don-t-use-glibc-specific-qsort_r.patch \
file://0003-missing_type.h-add-__compare_fn_t-and-comparison_fn_.patch \
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] Should systemd be marked as incompatible with musl?
On 5/23/19 6:45 PM, ChenQi wrote:
On 05/23/2019 08:22 PM, Burton, Ross wrote:
On Thu, 23 May 2019 at 11:34, Adrian Bunk wrote:
systemd is fundamentally Linux-only and not portable to other kernels.
systemd upstream is using glibc extensions not present in other
C libraries.
systemd upstream is accepting technically correct patches that help
building with musl, but there is no interest upstream in keeping systemd
working with non-glibc C libraries.
The way things currently go, systemd/musl will require an ever-growing
amount of not upstreamable patches - and this is not sustainable.
I think I have to agree with you: several of the extensions are for
security purposes, so we're potentially actively introducing issues.
Ross
I suggest we only apply these musl patches when the libc is musl.
SRC_URI += "${@d.getVar('SRC_URI_MUSL') if d.getVar('TCLIBC') == 'musl'
else ''}"
In this way, these patches do not affect glibc users. People who
actually use musl will test these patches and maybe improve them if they
really care about it.
thats acceptable approach. So far we have helped upstreams of apps to
cleanup or add support for musl and it has worked reasonably well,
systemd is a bit of odd man out since they explicitly seems to not care
for non glibc/linux systems, still they do accept cleanups which are
resulting from such ports so its not completely wasted excercise, but I
think dropping systemd support completely from musl is not an option I
would like to go with, there are cases where this makes sense.
Especially when you have to cater to different set of devices from small
to big, userspace remaining same is big advantage atleast in the world I
am in.
For example, the following patches are written by me.
0017-Do-not-disable-buffering-when-writing-to-oom_score_a.patch and
0001-do-not-disable-buffer-in-writing-files.patch
These two patches are to solve runtime problems in musl based system. I
knew there's something different in buffering mechanism between glibc
and musl that causes this problem, but I really did not want to spend
time digging into it. And as systemd was using unbuffered way for
writing files for a long time, I just disabled them.
As Adrian and Ross pointed out, these patches are introducing potential
security problems. I agree that we should do something about it. At a
minimum, the glibc based system should not be affected.
I'll send out patch and please help review it.
Best Regards,
Chen Qi
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH 1/1] systemd: avoid musl specific patches affect glibc systems
systemd upstream only care about glibc. We made musl specific
patches so that systemd could work. But currently these patches
contain potential security issues.
So apply these patches only when the libc is musl.
Signed-off-by: Chen Qi
---
meta/recipes-core/systemd/systemd_242.bb | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/meta/recipes-core/systemd/systemd_242.bb
b/meta/recipes-core/systemd/systemd_242.bb
index 2dda0d0..adb592f 100644
--- a/meta/recipes-core/systemd/systemd_242.bb
+++ b/meta/recipes-core/systemd/systemd_242.bb
@@ -27,7 +27,7 @@ SRC_URI += "file://touchscreen.rules \
"
# patches needed by musl
-SRC_URI += "${SRC_URI_MUSL}"
+SRC_URI += "${@d.getVar('SRC_URI_MUSL') if d.getVar('TCLIBC') == 'musl' else
''}"
SRC_URI_MUSL =
"file://0001-Use-getenv-when-secure-versions-are-not-available.patch \
file://0002-don-t-use-glibc-specific-qsort_r.patch \
file://0003-missing_type.h-add-__compare_fn_t-and-comparison_fn_.patch \
--
1.9.1
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH 0/1] systemd: avoid musl specific patches affect glibc systems
*** BLURB HERE *** The following changes since commit 06381a917e7e7648ace4746cbda55dcd138a64ec: perl/modules: Add various missing ptest perl module dependencies (2019-05-22 07:24:52 +0100) are available in the git repository at: git://git.pokylinux.org/poky-contrib ChenQi/systemd-musl http://git.pokylinux.org/cgit.cgi/poky-contrib/log/?h=ChenQi/systemd-musl Chen Qi (1): systemd: avoid musl specific patches affect glibc systems meta/recipes-core/systemd/systemd_242.bb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) -- 1.9.1 -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH] ofono: Add TEMP_FAILURE_RETRY optional definition
Fixes build with musl where TEMP_FAILURE_RETRY is not provided
Signed-off-by: Khem Raj
---
...tional-TEMP_FAILURE_RETRY-macro-copy.patch | 36 +++
meta/recipes-connectivity/ofono/ofono_1.29.bb | 1 +
2 files changed, 37 insertions(+)
create mode 100644
meta/recipes-connectivity/ofono/ofono/0001-mbim-add-an-optional-TEMP_FAILURE_RETRY-macro-copy.patch
diff --git
a/meta/recipes-connectivity/ofono/ofono/0001-mbim-add-an-optional-TEMP_FAILURE_RETRY-macro-copy.patch
b/meta/recipes-connectivity/ofono/ofono/0001-mbim-add-an-optional-TEMP_FAILURE_RETRY-macro-copy.patch
new file mode 100644
index 00..8a5a300adc
--- /dev/null
+++
b/meta/recipes-connectivity/ofono/ofono/0001-mbim-add-an-optional-TEMP_FAILURE_RETRY-macro-copy.patch
@@ -0,0 +1,36 @@
+From 22b52db4842611ac31a356f023fc09595384e2ad Mon Sep 17 00:00:00 2001
+From: Khem Raj
+Date: Thu, 23 May 2019 18:11:22 -0700
+Subject: [PATCH] mbim: add an optional TEMP_FAILURE_RETRY macro copy
+
+Fixes build on musl which does not provide this macro
+
+Upstream-Status: Submitted
[https://lists.ofono.org/pipermail/ofono/2019-May/019370.html]
+Signed-off-by: Khem Raj
+---
+ drivers/mbimmodem/mbim-private.h | 9 +
+ 1 file changed, 9 insertions(+)
+
+diff --git a/drivers/mbimmodem/mbim-private.h
b/drivers/mbimmodem/mbim-private.h
+index e159235..51693ea 100644
+--- a/drivers/mbimmodem/mbim-private.h
b/drivers/mbimmodem/mbim-private.h
+@@ -21,6 +21,15 @@
+
+ #define align_len(len, boundary) (((len)+(boundary)-1) & ~((boundary)-1))
+
++#ifndef TEMP_FAILURE_RETRY
++#define TEMP_FAILURE_RETRY(expression) ({ \
++ __typeof(expression) __result; \
++ do {\
++__result = (expression); \
++ } while (__result == -1 && errno == EINTR); \
++ __result; })
++#endif
++
+ enum mbim_control_message {
+ MBIM_OPEN_MSG = 0x1,
+ MBIM_CLOSE_MSG = 0x2,
+--
+2.21.0
+
diff --git a/meta/recipes-connectivity/ofono/ofono_1.29.bb
b/meta/recipes-connectivity/ofono/ofono_1.29.bb
index b5838e19bf..900a6d26d9 100644
--- a/meta/recipes-connectivity/ofono/ofono_1.29.bb
+++ b/meta/recipes-connectivity/ofono/ofono_1.29.bb
@@ -7,6 +7,7 @@ SRC_URI = "\
file://0001-build-Add-check-for-explicit_bzero-support.patch \
file://0001-main-Quiet-ld-errors-with-external-ell.patch \
file://0001-Makefile.am-Don-t-overwrite-src_ofonod_DEPENDENCIES.patch \
+ file://0001-mbim-add-an-optional-TEMP_FAILURE_RETRY-macro-copy.patch \
"
SRC_URI[md5sum] = "4fa0372630ff03f223452e4d05efa8f8"
SRC_URI[sha256sum] =
"67f0f8e5740dea5b46309e40667d1e560be39c90ef08dd01ff9e9ce8e61f0679"
--
2.21.0
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] connman: fix segfault with musl >2.21
On Thu, 2019-05-23 at 08:47 +0100, Nicola Lunghi wrote:
> musl > 2.21 changed the implementation of the freeaddrinfo() function
I think you mean 1.1.21 here and in the commit subject.
A.
> not allowing anymore to pass null pointers to it.
> This was causing a segmentation fault in connman.
> ---
> ...-gweb-fix-segfault-with-musl-v1.1.21.patch | 34 +++
> .../connman/connman_1.37.bb | 5 ++-
> 2 files changed, 38 insertions(+), 1 deletion(-)
> create mode 100644
> meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
>
> diff --git
> a/meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
> b/meta/recipes-
> connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
> new file mode 100644
> index 00..43b43bc9f8
> --- /dev/null
> +++
> b/meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
> @@ -0,0 +1,34 @@
> +From f0a8c69971b30ea7ca255bb885fdd1179fa5d298 Mon Sep 17 00:00:00 2001
> +From: Nicola Lunghi
> +Date: Thu, 23 May 2019 07:55:25 +0100
> +Subject: [PATCH] gweb: fix segfault with musl v1.1.21
> +
> +In musl > 1.1.21 freeaddrinfo() implementation changed and
> +was causing a segmentation fault on recent Yocto using musl.
> +
> +See this commit:
> +
> +
> https://git.musl-libc.org/cgit/musl/commit/src/network/freeaddrinfo.c?id=d1395c43c019aec6b855cf3c656bf47c8a719e7f
> +
> +Upstream-Status: Submitted
> +---
> + gweb/gweb.c | 3 ++-
> + 1 file changed, 2 insertions(+), 1 deletion(-)
> +
> +diff --git a/gweb/gweb.c b/gweb/gweb.c
> +index 393afe0a..12fcb1d8 100644
> +--- a/gweb/gweb.c
> b/gweb/gweb.c
> +@@ -1274,7 +1274,8 @@ static bool is_ip_address(const char *host)
> + addr = NULL;
> +
> + result = getaddrinfo(host, NULL, , );
> +-freeaddrinfo(addr);
> ++if(!result)
> ++freeaddrinfo(addr);
> +
> + return result == 0;
> + }
> +--
> +2.19.1
> +
> diff --git a/meta/recipes-connectivity/connman/connman_1.37.bb
> b/meta/recipes-connectivity/connman/connman_1.37.bb
> index 2cf904cd85..f52b21cae3 100644
> --- a/meta/recipes-connectivity/connman/connman_1.37.bb
> +++ b/meta/recipes-connectivity/connman/connman_1.37.bb
> @@ -7,7 +7,10 @@ SRC_URI =
> "${KERNELORG_MIRROR}/linux/network/${BPN}/${BP}.tar.xz \
> file://no-version-scripts.patch \
> "
>
> -SRC_URI_append_libc-musl = "
> file://0002-resolve-musl-does-not-implement-res_ninit.patch"
> +SRC_URI_append_libc-musl = " \
> +file://0002-resolve-musl-does-not-implement-res_ninit.patch \
> +file://0003-gweb-fix-segfault-with-musl-v1.1.21.patch \
> +"
>
> SRC_URI[md5sum] = "75012084f14fb63a84b116e66c6e94fb"
> SRC_URI[sha256sum] =
> "6ce29b3eb0bb16a7387bc609c39455fd13064bdcde5a4d185fab3a0c71946e16"
> --
> 2.19.1
>
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] Should systemd be marked as incompatible with musl?
systemd is fundamentally Linux-only and not portable to other kernels. systemd upstream is using glibc extensions not present in other C libraries. systemd upstream is accepting technically correct patches that help building with musl, but there is no interest upstream in keeping systemd working with non-glibc C libraries. The way things currently go, systemd/musl will require an ever-growing amount of not upstreamable patches - and this is not sustainable. It is also not clear where the systemd/musl combination makes sense: If an embedded system is so size limited that the size of glibc matters, then systemd is unlikely to be an option. Other musl-using distribution also seem to favour more lightweight init systems. If someone has important usecases for using systemd with musl it might be possible to solve find solutions in an upstreamable way, but this would require a significant longterm commitment. The current state is rather bad, some examples: 0001-Use-getenv-when-secure-versions-are-not-available.patch looks like waiting for a CVE number. 0008-add-missing-FTW_-macros-for-musl.patch seems to break code in areas (rng, smack, selinux) where breakages can have security implications. 0002-don-t-use-glibc-specific-qsort_r.patch might introduce race conditions that cause data corruption. 0007-don-t-fail-if-GLOB_BRACE-and-GLOB_ALTDIRFUNC-is-not.patch makes some functions behave different from what is expected by callers. 0017-Do-not-disable-buffering-when-writing-to-oom_score_a.patch and 0001-do-not-disable-buffer-in-writing-files.patch seems like an issue that would require debugging the root cause and then addressing it. All this looks bad, and is expected to become worse. In this situation it would be better to not claim that using systemd with musl would be a supported option. cu Adrian -- "Is there not promise of rain?" Ling Tan asked suddenly out of the darkness. There had been need of rain for many days. "Only a promise," Lao Er said. Pearl S. Buck - Dragon Seed -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH] chkconfig: do not use unstable github archive tarballs
Signed-off-by: Alexander Kanavin
---
meta/recipes-extended/chkconfig/chkconfig_1.3.58.bb | 9 -
1 file changed, 4 insertions(+), 5 deletions(-)
diff --git a/meta/recipes-extended/chkconfig/chkconfig_1.3.58.bb
b/meta/recipes-extended/chkconfig/chkconfig_1.3.58.bb
index 2f1f6c02640..788a37b6ed0 100644
--- a/meta/recipes-extended/chkconfig/chkconfig_1.3.58.bb
+++ b/meta/recipes-extended/chkconfig/chkconfig_1.3.58.bb
@@ -18,14 +18,13 @@ PR = "r7"
S = "${WORKDIR}/${BPN}-${BPN}-${PV}"
-UPSTREAM_CHECK_URI = "https://github.com/fedora-sysv/${BPN}/releases;
-
-SRC_URI =
"https://github.com/fedora-sysv/chkconfig/archive/chkconfig-${PV}.tar.gz \
+SRC_URI = "git://github.com/fedora-sysv/chkconfig.git \
file://replace_caddr_t.patch \
"
-SRC_URI[md5sum] = "3f51ac38a234be5278b3a2d9705eda5e"
-SRC_URI[sha256sum] =
"bf1e81f0d7cc999b536c9fe7877abf584a4082fd03c9d2597b6f090966579b40"
+SRCREV = "3e0b3909ff6328d232b27b9c32f1a9309c6cd0a4"
+
+S = "${WORKDIR}/git"
inherit gettext
--
2.17.1
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH] chkconfig: fix upstream version check
Signed-off-by: Alexander Kanavin
---
meta/recipes-extended/chkconfig/chkconfig_1.3.58.bb | 1 +
1 file changed, 1 insertion(+)
diff --git a/meta/recipes-extended/chkconfig/chkconfig_1.3.58.bb
b/meta/recipes-extended/chkconfig/chkconfig_1.3.58.bb
index 788a37b6ed0..d21dd697042 100644
--- a/meta/recipes-extended/chkconfig/chkconfig_1.3.58.bb
+++ b/meta/recipes-extended/chkconfig/chkconfig_1.3.58.bb
@@ -23,6 +23,7 @@ SRC_URI = "git://github.com/fedora-sysv/chkconfig.git \
"
SRCREV = "3e0b3909ff6328d232b27b9c32f1a9309c6cd0a4"
+UPSTREAM_CHECK_GITTAGREGEX = "(?P\d+(\.\d+)+)"
S = "${WORKDIR}/git"
--
2.17.1
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH v3 0/2] Make systemd as default init manager and configure wired network
From: Kai Kang
v3:
* add the patch to make systemd as default init manager
* rename the .network with wired.network
* call to bb.utils.contains('OVERRIDES', ...) is not right. Install
wired.network unconditionally and remove in in do_install_append_qemuall()
BTW, bb.utils.contains('IMAGE_CLASSES', 'qemuboot', ...) from layer
meta-yoe doesn't work for genericx86 that why I didn't use it.
Testededed on genericx86-64.
Run testimage with qemux86-64:
MACHINE ?= "qemux86-64"
IMAGE_CLASSES += "testimage"
QEMU_USE_SLIRP = "1"
TEST_SERVER_IP = "xxx"
$ bitbake core-image-sato -c testimage
SUMMARY:
core-image-sato () - Ran 60 tests in 1042.949s
core-image-sato - OK - All required tests passed (successes=42, skipped=18,
failures=0, errors=0)
NOTE: Tasks Summary: Attempted 723 tasks of which 722 didn't need to be rerun
and all succeeded.
Kai Kang (2):
local.conf.sample: make systemd as default init manager
systemd-conf: configure wired network with dhcp
meta-poky/conf/local.conf.sample | 9 +
meta-poky/conf/local.conf.sample.extended| 9 -
meta/recipes-core/systemd/systemd-conf/wired.network | 9 +
meta/recipes-core/systemd/systemd-conf_242.bb| 6 ++
4 files changed, 24 insertions(+), 9 deletions(-)
create mode 100644 meta/recipes-core/systemd/systemd-conf/wired.network
--
2.18.0
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH v3 2/2] systemd-conf: configure wired network with dhcp
From: Kai Kang
Add a configure file for systemd.networkd to configure wired network
interfaces with dhcp. It works with common network interfaces such eth0
and eno1. And do not install it for qemu bsps.
Refer to
https://github.com/YoeDistro/meta-yoe/tree/master/recipes-core/systemd
[YOCTO #13057]
Signed-off-by: Kai Kang
---
meta/recipes-core/systemd/systemd-conf/wired.network | 9 +
meta/recipes-core/systemd/systemd-conf_242.bb| 6 ++
2 files changed, 15 insertions(+)
create mode 100644 meta/recipes-core/systemd/systemd-conf/wired.network
diff --git a/meta/recipes-core/systemd/systemd-conf/wired.network
b/meta/recipes-core/systemd/systemd-conf/wired.network
new file mode 100644
index 00..253aee9382
--- /dev/null
+++ b/meta/recipes-core/systemd/systemd-conf/wired.network
@@ -0,0 +1,9 @@
+[Match]
+Name=en* eth*
+
+[Network]
+DHCP=yes
+
+[DHCP]
+RouteMetric=10
+ClientIdentifier=mac
diff --git a/meta/recipes-core/systemd/systemd-conf_242.bb
b/meta/recipes-core/systemd/systemd-conf_242.bb
index 96beea53a7..d9ec023bfd 100644
--- a/meta/recipes-core/systemd/systemd-conf_242.bb
+++ b/meta/recipes-core/systemd/systemd-conf_242.bb
@@ -10,17 +10,22 @@ SRC_URI = "\
file://logind.conf \
file://system.conf \
file://system.conf-qemuall \
+file://wired.network \
"
do_install() {
install -D -m0644 ${WORKDIR}/journald.conf
${D}${systemd_unitdir}/journald.conf.d/00-${PN}.conf
install -D -m0644 ${WORKDIR}/logind.conf
${D}${systemd_unitdir}/logind.conf.d/00-${PN}.conf
install -D -m0644 ${WORKDIR}/system.conf
${D}${systemd_unitdir}/system.conf.d/00-${PN}.conf
+ install -D -m0644 ${WORKDIR}/wired.network
${D}${systemd_unitdir}/network/80-wired.network
}
# Based on change from YP bug 8141, OE commit
5196d7bacaef1076c361adaa2867be31759c1b52
do_install_append_qemuall() {
install -D -m0644 ${WORKDIR}/system.conf-qemuall
${D}${systemd_unitdir}/system.conf.d/01-${PN}.conf
+
+ # Do not install wired.network for qemu bsps
+ rm -rf ${D}${systemd_unitdir}/network
}
PACKAGE_ARCH = "${MACHINE_ARCH}"
@@ -29,4 +34,5 @@ FILES_${PN} = "\
${systemd_unitdir}/journald.conf.d/ \
${systemd_unitdir}/logind.conf.d/ \
${systemd_unitdir}/system.conf.d/ \
+${systemd_unitdir}/network/ \
"
--
2.18.0
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH] connman: fix segfault with musl >2.21
On Thu, 2019-05-23 at 08:47 +0100, Nicola Lunghi wrote:
> musl > 2.21 changed the implementation of the freeaddrinfo() function
> not allowing anymore to pass null pointers to it.
> This was causing a segmentation fault in connman.
> ---
> ...-gweb-fix-segfault-with-musl-v1.1.21.patch | 34 +++
> .../connman/connman_1.37.bb | 5 ++-
> 2 files changed, 38 insertions(+), 1 deletion(-)
> create mode 100644
> meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
>
> diff --git
> a/meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
> b/meta/recipes-
> connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
> new file mode 100644
> index 00..43b43bc9f8
> --- /dev/null
> +++
> b/meta/recipes-connectivity/connman/connman/0003-gweb-fix-segfault-with-musl-v1.1.21.patch
> @@ -0,0 +1,34 @@
> +From f0a8c69971b30ea7ca255bb885fdd1179fa5d298 Mon Sep 17 00:00:00 2001
> +From: Nicola Lunghi
> +Date: Thu, 23 May 2019 07:55:25 +0100
> +Subject: [PATCH] gweb: fix segfault with musl v1.1.21
> +
> +In musl > 1.1.21 freeaddrinfo() implementation changed and
> +was causing a segmentation fault on recent Yocto using musl.
> +
> +See this commit:
> +
> +
> https://git.musl-libc.org/cgit/musl/commit/src/network/freeaddrinfo.c?id=d1395c43c019aec6b855cf3c656bf47c8a719e7f
> +
> +Upstream-Status: Submitted
> +---
> + gweb/gweb.c | 3 ++-
> + 1 file changed, 2 insertions(+), 1 deletion(-)
> +
> +diff --git a/gweb/gweb.c b/gweb/gweb.c
> +index 393afe0a..12fcb1d8 100644
> +--- a/gweb/gweb.c
> b/gweb/gweb.c
> +@@ -1274,7 +1274,8 @@ static bool is_ip_address(const char *host)
> + addr = NULL;
> +
> + result = getaddrinfo(host, NULL, , );
> +-freeaddrinfo(addr);
> ++if(!result)
> ++freeaddrinfo(addr);
> +
> + return result == 0;
> + }
> +--
> +2.19.1
> +
> diff --git a/meta/recipes-connectivity/connman/connman_1.37.bb
> b/meta/recipes-connectivity/connman/connman_1.37.bb
> index 2cf904cd85..f52b21cae3 100644
> --- a/meta/recipes-connectivity/connman/connman_1.37.bb
> +++ b/meta/recipes-connectivity/connman/connman_1.37.bb
> @@ -7,7 +7,10 @@ SRC_URI =
> "${KERNELORG_MIRROR}/linux/network/${BPN}/${BP}.tar.xz \
> file://no-version-scripts.patch \
> "
>
> -SRC_URI_append_libc-musl = "
> file://0002-resolve-musl-does-not-implement-res_ninit.patch"
> +SRC_URI_append_libc-musl = " \
> +file://0002-resolve-musl-does-not-implement-res_ninit.patch \
> +file://0003-gweb-fix-segfault-with-musl-v1.1.21.patch \
> +"
This fix is not musl specific.
Cheers,
Andre'
>
> SRC_URI[md5sum] = "75012084f14fb63a84b116e66c6e94fb"
> SRC_URI[sha256sum] =
> "6ce29b3eb0bb16a7387bc609c39455fd13064bdcde5a4d185fab3a0c71946e16"
> --
> 2.19.1
>
--
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core
[OE-core] [PATCH v3 1/2] local.conf.sample: make systemd as default init manager
From: Kai Kang Move configurations from local.conf.sample.extended to local.conf.sample to make systemd as default init manager for poky. [YOCTO #13031] Signed-off-by: Kai Kang --- meta-poky/conf/local.conf.sample | 9 + meta-poky/conf/local.conf.sample.extended | 9 - 2 files changed, 9 insertions(+), 9 deletions(-) diff --git a/meta-poky/conf/local.conf.sample b/meta-poky/conf/local.conf.sample index 9068e567dc..3a07105e44 100644 --- a/meta-poky/conf/local.conf.sample +++ b/meta-poky/conf/local.conf.sample @@ -249,3 +249,12 @@ PACKAGECONFIG_append_pn-nativesdk-qemu = " sdl" # track the version of this file when it was generated. This can safely be ignored if # this doesn't mean anything to you. CONF_VERSION = "1" + +# +# Use systemd for system initialization +# +DISTRO_FEATURES_append = " systemd" +DISTRO_FEATURES_BACKFILL_CONSIDERED += "sysvinit" +VIRTUAL-RUNTIME_login_manager = "shadow-base" +VIRTUAL-RUNTIME_init_manager = "systemd" +VIRTUAL-RUNTIME_initscripts = "systemd-compat-units" diff --git a/meta-poky/conf/local.conf.sample.extended b/meta-poky/conf/local.conf.sample.extended index 26603debe6..e20fc5dbf3 100644 --- a/meta-poky/conf/local.conf.sample.extended +++ b/meta-poky/conf/local.conf.sample.extended @@ -377,12 +377,3 @@ DISTRO_FEATURES_remove = "x11" #VIRTUAL-RUNTIME_initscripts = "initscripts" #VIRTUAL-RUNTIME_keymaps = "keymaps" #DISTRO_FEATURES_BACKFILL_CONSIDERED += "sysvinit" - -# -# Use systemd for system initialization -# -#DISTRO_FEATURES_append = " systemd" -#DISTRO_FEATURES_BACKFILL_CONSIDERED += "sysvinit" -#VIRTUAL-RUNTIME_login_manager = "shadow-base" -#VIRTUAL-RUNTIME_init_manager = "systemd" -#VIRTUAL-RUNTIME_initscripts = "systemd-compat-units" -- 2.18.0 -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
Re: [OE-core] [PATCH 1/2] piglit: upgrade to latest revision
Yes. If devtool drops them, then they are already in the code.
Alex
On Thu, 23 May 2019 at 03:30, Khem Raj wrote:
>
> On Wed, May 22, 2019 at 4:02 AM Alexander Kanavin
> wrote:
> >
> > Signed-off-by: Alexander Kanavin
> > ---
> > ...UPWARD-only-if-its-defined-in-fenv.h.patch | 75 ---
> > .../piglit/piglit/format-fix.patch| 69 -
>
> are these patches upstreamed ?
>
> > meta/recipes-graphics/piglit/piglit_git.bb| 5 +-
> > 3 files changed, 1 insertion(+), 148 deletions(-)
> > delete mode 100644
> > meta/recipes-graphics/piglit/piglit/0001-tests-Use-FE_UPWARD-only-if-its-defined-in-fenv.h.patch
> > delete mode 100644 meta/recipes-graphics/piglit/piglit/format-fix.patch
> >
> > diff --git
> > a/meta/recipes-graphics/piglit/piglit/0001-tests-Use-FE_UPWARD-only-if-its-defined-in-fenv.h.patch
> >
> > b/meta/recipes-graphics/piglit/piglit/0001-tests-Use-FE_UPWARD-only-if-its-defined-in-fenv.h.patch
> > deleted file mode 100644
> > index 57eda2e207b..000
> > ---
> > a/meta/recipes-graphics/piglit/piglit/0001-tests-Use-FE_UPWARD-only-if-its-defined-in-fenv.h.patch
> > +++ /dev/null
> > @@ -1,75 +0,0 @@
> > -Upstream-Status: Submitted [mailing list]
> > -Signed-off-by: Ross Burton
> > -
> > -From 0e0a2a69261031d55d52b6045990e8982ea12912 Mon Sep 17 00:00:00 2001
> > -From: Khem Raj
> > -Date: Sat, 9 Jul 2016 07:52:19 +
> > -Subject: [PATCH] tests: only run rounding tests if FE_UPWARD is present
> > -
> > -On ARM, musl does not define FE_* when the architecture does not have VFP
> > (which
> > -is the right interpretation).
> > -
> > -As these tests depend on calling fesetround(), skip the test if FE_UPWARD
> > isn't
> > -available.
> > -
> > -Signed-off-by: Ross Burton
> >
> > - tests/general/roundmode-getintegerv.c | 12
> > - tests/general/roundmode-pixelstore.c | 12
> > - 2 files changed, 16 insertions(+), 8 deletions(-)
> > -
> > -diff --git a/tests/general/roundmode-getintegerv.c
> > b/tests/general/roundmode-getintegerv.c
> > -index 28ecfaf55..aa99044a1 100644
> > a/tests/general/roundmode-getintegerv.c
> > -+++ b/tests/general/roundmode-getintegerv.c
> > -@@ -79,13 +79,17 @@ test(float val, int expect)
> > - void
> > - piglit_init(int argc, char **argv)
> > - {
> > -- int ret;
> > - bool pass = true;
> > -- ret = fesetround(FE_UPWARD);
> > -- if (ret != 0) {
> > -- printf("Couldn't set rounding mode\n");
> > -+
> > -+#ifdef FE_UPWARD
> > -+ if (fesetround(FE_UPWARD) != 0) {
> > -+ printf("Setting rounding mode failed\n");
> > - piglit_report_result(PIGLIT_SKIP);
> > - }
> > -+#else
> > -+ printf("Cannot set rounding mode\n");
> > -+ piglit_report_result(PIGLIT_SKIP);
> > -+#endif
> > -
> > - pass = test(2.2, 2) && pass;
> > - pass = test(2.8, 3) && pass;
> > -diff --git a/tests/general/roundmode-pixelstore.c
> > b/tests/general/roundmode-pixelstore.c
> > -index 8a029b257..57ec11c09 100644
> > a/tests/general/roundmode-pixelstore.c
> > -+++ b/tests/general/roundmode-pixelstore.c
> > -@@ -79,13 +79,17 @@ test(float val, int expect)
> > - void
> > - piglit_init(int argc, char **argv)
> > - {
> > -- int ret;
> > - bool pass = true;
> > -- ret = fesetround(FE_UPWARD);
> > -- if (ret != 0) {
> > -- printf("Couldn't set rounding mode\n");
> > -+
> > -+#ifdef FE_UPWARD
> > -+ if (fesetround(FE_UPWARD) != 0) {
> > -+ printf("Setting rounding mode failed\n");
> > - piglit_report_result(PIGLIT_SKIP);
> > - }
> > -+#else
> > -+ printf("Cannot set rounding mode\n");
> > -+ piglit_report_result(PIGLIT_SKIP);
> > -+#endif
> > -
> > - pass = test(2.2, 2) && pass;
> > - pass = test(2.8, 3) && pass;
> > ---
> > -2.11.0
> > -
> > diff --git a/meta/recipes-graphics/piglit/piglit/format-fix.patch
> > b/meta/recipes-graphics/piglit/piglit/format-fix.patch
> > deleted file mode 100644
> > index 73d539fef27..000
> > --- a/meta/recipes-graphics/piglit/piglit/format-fix.patch
> > +++ /dev/null
> > @@ -1,69 +0,0 @@
> > -Upstream-Status: Submitted [mailing list]
> > -Signed-off-by: Ross Burton
> > -
> > -From f0c6981322807e179e39ce67aeebd42cf7a54d36 Mon Sep 17 00:00:00 2001
> > -From: Ross Burton
> > -Date: Wed, 21 Nov 2018 12:44:36 +
> > -Subject: [PATCH] arb_texture_view: fix security format warnings
> > -
> > -If built with -Werror=format-security then Piglit fails to build:
> > -
> > -/tests/spec/arb_texture_view/rendering-layers-image.c:150:8:
> > -error: format not a string literal and no format arguments
> > [-Werror=format-security]
> > - (desc)); \
> > - ^~
> > -
> > -In this case test->uniform_type is being turned into a string using
> > snprintf()
> > -and then passed to piglit_report_subtest_result() which takes a format
> > string,
> > -but GCC can't verify the format.
> > -
> >
Re: [OE-core] [PATCH v2] insane: add sanity checks to SRC_URI
On Thu, 23 May 2019 at 00:24, Richard Purdie wrote: > Ultimately I'd say that all these checks should work with INSANE_SKIP. > I was just wondering about this issue myself for the chkconfig > autobuilder warnings... chkconfig can be easily addressed by switching to git:// I think. Patch coming. Alex -- ___ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core
