Skip to contentSkip to conversation

package/libnss: use correct CFLAGS and LDFLAGS

PATCH Superseded Giulio Benetti 3 participants 6 messages Jul 29, 2019 – Aug 20, 2019

lore
Share & tools
lore
#0 lore raw

PATCH v3 1/1 package/libnss: use correct CFLAGS and LDFLAGS

Actually libnss is configure with BUILD_OPT=1 that doesn't take into account TARGET_CFLAGS nor TARGET_LDFLAGS resulting in pre-chosen optimizations and debugging symbols.

Can't pass TARGET_CFLAGS through CFLAGS or similar otherwise other internal libnss one will be overwritten(i.e. -fPIC), so prefer to append TARGET_CFLAGS at the end of Linux.mk as well as TARGET_LDFLAGS according to internal libnss Makefile system's names. And obviously remove BUILD_OPT=1 from BUILD_VARS. This reveals hidden bugs when building with -Os due to uninitialized variables so add patches to fix those bugs.

---
V1->V2:
* use $() instead of ${}
* add patches to fix all warnings when building with -Os or -Og V2->V3:
* substitute all patches with upstreamed ones

---
 ...ve-Wmaybe-uninitialized-warning-in-t.patch | 30 +++++++++++
 ...ve-Wmaybe-uninitialized-warning-in-h.patch | 30 +++++++++++
 ...ve-Wmaybe-uninitialized-warning-in-l.patch | 30 +++++++++++
 ...ve-Wmaybe-uninitialized-warning-in-t.patch | 30 +++++++++++
 ...ve-Wmaybe-uninitialized-warning-in-p.patch | 40 +++++++++++++++
 ...ve-Wmaybe-uninitialized-warning-in-p.patch | 39 +++++++++++++++
 ...ve-Wmaybe-uninitialized-warning-in-p.patch | 50 +++++++++++++++++++
 package/libnss/libnss.mk                      | 10 +++-
 8 files changed, 257 insertions(+), 2 deletions(-)
 create mode 100644 package/libnss/0003-Bug-1561556-Remove-Wmaybe-uninitialized-warning-in-t.patch
 create mode 100644 package/libnss/0004-Bug-1561558-Remove-Wmaybe-uninitialized-warning-in-h.patch
 create mode 100644 package/libnss/0005-Bug-1561587-Remove-Wmaybe-uninitialized-warning-in-l.patch
 create mode 100644 package/libnss/0006-Bug-1561591-Remove-Wmaybe-uninitialized-warning-in-t.patch
 create mode 100644 package/libnss/0007-Bug-1561548-Remove-Wmaybe-uninitialized-warning-in-p.patch
 create mode 100644 package/libnss/0008-Bug-1561588-Remove-Wmaybe-uninitialized-warning-in-p.patch
 create mode 100644 package/libnss/0009-Bug-1561598-Remove-Wmaybe-uninitialized-warning-in-p.patch
package/libnss/0003-Bug-1561556-Remove-Wmaybe-uninitialized-warning-in-t.patch new file +30 -0
new file mode 100644
index 0000000000..bfae2206cb
@@ -0,0 +1,30 @@
+From 659daaa7e09c2a978cd71407d65bf8e1839104b7 Mon Sep 17 00:00:00 2001
+From: Giulio Benetti <[email protected]>
+Date: Wed, 10 Jul 2019 14:58:32 -0700
+Subject: [PATCH 1/7] Bug 1561556 - Remove -Wmaybe-uninitialized warning in
+ tls13esni.c r=marcusburghardt,jcj
+
+--HG--
+extra : source : 9ab49a84249573e6615e02f6b7f3060b925c1b8b
+
+Signed-off-by: Giulio Benetti <[email protected]>
+---
+ nss/lib/ssl/tls13esni.c | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/nss/lib/ssl/tls13esni.c b/nss/lib/ssl/tls13esni.c
+index 9b635a9cf..4d2e12d62 100644
+--- a/nss/lib/ssl/tls13esni.c
++++ b/nss/lib/ssl/tls13esni.c
+@@ -580,7 +580,7 @@ tls13_ClientSetupESNI(sslSocket *ss)
+     size_t i;
+     PRCList *cur;
+     SECStatus rv;
+-    TLS13KeyShareEntry *share;
++    TLS13KeyShareEntry *share = NULL;
+     const sslNamedGroupDef *group = NULL;
+     PRTime now = PR_Now() / PR_USEC_PER_SEC;
+ 
+-- 
+2.17.1
+
package/libnss/0004-Bug-1561558-Remove-Wmaybe-uninitialized-warning-in-h.patch new file +30 -0
new file mode 100644
index 0000000000..6e40d12242
@@ -0,0 +1,30 @@
+From 453d97dff76635bfb4aba4a8f3626c1536ec07e7 Mon Sep 17 00:00:00 2001
+From: Giulio Benetti <[email protected]>
+Date: Wed, 10 Jul 2019 15:00:24 -0700
+Subject: [PATCH 2/7] Bug 1561558 - Remove -Wmaybe-uninitialized warning in
+ httpserv.c r=marcusburghardt,jcj
+
+--HG--
+extra : source : 2f3d66e0b4142e4cacf3fe11267176f0371c1b48
+
+Signed-off-by: Giulio Benetti <[email protected]>
+---
+ nss/cmd/httpserv/httpserv.c | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/nss/cmd/httpserv/httpserv.c b/nss/cmd/httpserv/httpserv.c
+index 71e2ab88d..c7277f3bd 100644
+--- a/nss/cmd/httpserv/httpserv.c
++++ b/nss/cmd/httpserv/httpserv.c
+@@ -463,7 +463,7 @@ handle_connection(
+     char *getData = NULL; /* inplace conversion */
+     SECItem postData;
+     PRBool isOcspRequest = PR_FALSE;
+-    PRBool isPost;
++    PRBool isPost = PR_FALSE;
+ 
+     postData.data = NULL;
+     postData.len = 0;
+-- 
+2.17.1
+
package/libnss/0005-Bug-1561587-Remove-Wmaybe-uninitialized-warning-in-l.patch new file +30 -0
new file mode 100644
index 0000000000..0681ee48d1
@@ -0,0 +1,30 @@
+From 6bedef511379d82babee408242aadb162eb732d6 Mon Sep 17 00:00:00 2001
+From: Giulio Benetti <[email protected]>
+Date: Wed, 10 Jul 2019 15:01:18 -0700
+Subject: [PATCH 3/7] Bug 1561587 - Remove -Wmaybe-uninitialized warning in
+ lgattr.c r=marcusburghardt,jcj
+
+--HG--
+extra : source : 346a2349d14872a63ca5e849750bfc6342a66a52
+
+Signed-off-by: Giulio Benetti <[email protected]>
+---
+ nss/lib/softoken/legacydb/lgattr.c | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/nss/lib/softoken/legacydb/lgattr.c b/nss/lib/softoken/legacydb/lgattr.c
+index 3d77bd056..c1865a38e 100644
+--- a/nss/lib/softoken/legacydb/lgattr.c
++++ b/nss/lib/softoken/legacydb/lgattr.c
+@@ -1069,7 +1069,7 @@ lg_FindTrustAttribute(LGObjectCache *obj, CK_ATTRIBUTE_TYPE type,
+     NSSLOWCERTCertificate *cert;
+     unsigned char hash[SHA1_LENGTH];
+     unsigned int trustFlags;
+-    CK_RV crv;
++    CK_RV crv = CKR_CANCEL;
+ 
+     switch (type) {
+         case CKA_PRIVATE:
+-- 
+2.17.1
+
package/libnss/0006-Bug-1561591-Remove-Wmaybe-uninitialized-warning-in-t.patch new file +30 -0
new file mode 100644
index 0000000000..bb9c4e9216
@@ -0,0 +1,30 @@
+From 871c12951e98d8a098ea7dcc26e249553c6633fb Mon Sep 17 00:00:00 2001
+From: Giulio Benetti <[email protected]>
+Date: Wed, 10 Jul 2019 15:01:49 -0700
+Subject: [PATCH 4/7] Bug 1561591 - Remove -Wmaybe-uninitialized warning in
+ tstclnt.c r=marcusburghardt,jcj
+
+--HG--
+extra : source : 4b09a45373c66567d8da9918f9ecaa053ebb10c7
+
+Signed-off-by: Giulio Benetti <[email protected]>
+---
+ nss/cmd/tstclnt/tstclnt.c | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/nss/cmd/tstclnt/tstclnt.c b/nss/cmd/tstclnt/tstclnt.c
+index 12c6df045..bb626d903 100644
+--- a/nss/cmd/tstclnt/tstclnt.c
++++ b/nss/cmd/tstclnt/tstclnt.c
+@@ -924,7 +924,7 @@ restartHandshakeAfterServerCertIfNeeded(PRFileDesc *fd,
+                                         PRBool override)
+ {
+     SECStatus rv;
+-    PRErrorCode error;
++    PRErrorCode error = 0;
+ 
+     if (!serverCertAuth->isPaused)
+         return SECSuccess;
+-- 
+2.17.1
+
package/libnss/0007-Bug-1561548-Remove-Wmaybe-uninitialized-warning-in-p.patch new file +40 -0
new file mode 100644
index 0000000000..e49bda6cd6
@@ -0,0 +1,40 @@
+From 09e48ebe2f29686c0f5ba2000303c6c88a5fd72d Mon Sep 17 00:00:00 2001
+From: Giulio Benetti <[email protected]>
+Date: Fri, 19 Jul 2019 09:39:39 -0700
+Subject: [PATCH 5/7] Bug 1561548 - Remove -Wmaybe-uninitialized warning in
+ pkix_pl_ldapdefaultclient.c r=marcusburghardt,jcj
+
+--HG--
+extra : amend_source : 66d4681f81ad23b26160d286a4d12139538ac1a4
+
+Signed-off-by: Giulio Benetti <[email protected]>
+---
+ nss/lib/libpkix/pkix_pl_nss/module/pkix_pl_ldapdefaultclient.c | 5 +++--
+ 1 file changed, 3 insertions(+), 2 deletions(-)
+
+diff --git a/nss/lib/libpkix/pkix_pl_nss/module/pkix_pl_ldapdefaultclient.c b/nss/lib/libpkix/pkix_pl_nss/module/pkix_pl_ldapdefaultclient.c
+index 3dc06be9a..9b6f8d688 100644
+--- a/nss/lib/libpkix/pkix_pl_nss/module/pkix_pl_ldapdefaultclient.c
++++ b/nss/lib/libpkix/pkix_pl_nss/module/pkix_pl_ldapdefaultclient.c
+@@ -352,7 +352,9 @@ pkix_pl_LdapDefaultClient_VerifyBindResponse(
+         SECItem decode = {siBuffer, NULL, 0};
+         SECStatus rv = SECFailure;
+         LDAPMessage msg;
+-        LDAPBindResponse *ldapBindResponse = NULL;
++        LDAPBindResponse *ldapBindResponse = &msg.protocolOp.op.bindResponseMsg;
++
++        ldapBindResponse->resultCode.data = NULL;
+ 
+         PKIX_ENTER
+                 (LDAPDEFAULTCLIENT,
+@@ -367,7 +369,6 @@ pkix_pl_LdapDefaultClient_VerifyBindResponse(
+                 PKIX_LDAPDEFAULTCLIENTDECODEBINDRESPONSEFAILED);
+ 
+         if (rv == SECSuccess) {
+-                ldapBindResponse = &msg.protocolOp.op.bindResponseMsg;
+                 if (*(ldapBindResponse->resultCode.data) == SUCCESS) {
+                         client->connectStatus = BOUND;
+                 } else {
+-- 
+2.17.1
+
package/libnss/0008-Bug-1561588-Remove-Wmaybe-uninitialized-warning-in-p.patch new file +39 -0
new file mode 100644
index 0000000000..782b80aa86
@@ -0,0 +1,39 @@
+From f8379a4edd0a8d5b6d7a411956b789785979e4c9 Mon Sep 17 00:00:00 2001
+From: Giulio Benetti <[email protected]>
+Date: Fri, 19 Jul 2019 09:40:51 -0700
+Subject: [PATCH 6/7] Bug 1561588 - Remove -Wmaybe-uninitialized warning in
+ p7env.c r=marcusburghardt,jcj
+
+--HG--
+extra : amend_source : 6acce3136f3b96899e8daca8df50d418d4ea0e02
+
+Signed-off-by: Giulio Benetti <[email protected]>
+---
+ nss/cmd/p7env/p7env.c | 4 ++--
+ 1 file changed, 2 insertions(+), 2 deletions(-)
+
+diff --git a/nss/cmd/p7env/p7env.c b/nss/cmd/p7env/p7env.c
+index b35bf9a98..453cbee19 100644
+--- a/nss/cmd/p7env/p7env.c
++++ b/nss/cmd/p7env/p7env.c
+@@ -66,7 +66,7 @@ EncryptFile(FILE *outFile, FILE *inFile, struct recipient *recipients,
+     SEC_PKCS7ContentInfo *cinfo;
+     SEC_PKCS7EncoderContext *ecx;
+     struct recipient *rcpt;
+-    SECStatus rv;
++    SECStatus rv = SECFailure;
+ 
+     if (outFile == NULL || inFile == NULL || recipients == NULL)
+         return -1;
+@@ -133,7 +133,7 @@ main(int argc, char **argv)
+     struct recipient *recipients, *rcpt;
+     PLOptState *optstate;
+     PLOptStatus status;
+-    SECStatus rv;
++    SECStatus rv = SECFailure;
+ 
+     progName = strrchr(argv[0], '/');
+     progName = progName ? progName + 1 : argv[0];
+-- 
+2.17.1
+
package/libnss/0009-Bug-1561598-Remove-Wmaybe-uninitialized-warning-in-p.patch new file +50 -0

50 lines not on this page · raw message

package/libnss/libnss.mk +8 -2
index 89d614fe1c..e1b9e02a77 100644
@@ -13,6 +13,14 @@ LIBNSS_DEPENDENCIES = libnspr sqlite zlib
 LIBNSS_LICENSE = MPL-2.0
 LIBNSS_LICENSE_FILES = nss/COPYING
 
+# Need to pass down TARGET_CFLAGS and TARGET_LDFLAGS
+define LIBNSS_FIXUP_LINUX_MK
+	echo 'OS_CFLAGS += $(TARGET_CFLAGS)' >> $(@D)/nss/coreconf/Linux.mk
+	echo 'LDFLAGS += $(TARGET_LDFLAGS)' >> $(@D)/nss/coreconf/Linux.mk
+endef
+
+LIBNSS_PRE_CONFIGURE_HOOKS += LIBNSS_FIXUP_LINUX_MK
+
 # --gc-sections triggers binutils ld segfault
 # https://sourceware.org/bugzilla/show_bug.cgi?id=21180
 ifeq ($(BR2_microblaze),y)
@@ -33,7 +41,6 @@ LIBNSS_BUILD_VARS = \
 	MOZILLA_CLIENT=1 \
 	NSPR_INCLUDE_DIR=$(STAGING_DIR)/usr/include/nspr \
 	NSPR_LIB_DIR=$(STAGING_DIR)/usr/lib \
-	BUILD_OPT=1 \
 	NS_USE_GCC=1 \
 	NSS_DISABLE_GTESTS=1 \
 	NSS_USE_SYSTEM_SQLITE=1 \
@@ -102,7 +109,6 @@ HOST_LIBNSS_BUILD_VARS = \
 	MOZILLA_CLIENT=1 \
 	NSPR_INCLUDE_DIR=$(HOST_DIR)/include/nspr \
 	NSPR_LIB_DIR=$(HOST_DIR)/lib \
-	BUILD_OPT=1 \
 	NS_USE_GCC=1 \
 	NSS_DISABLE_GTESTS=1 \
 	NSS_USE_SYSTEM_SQLITE=1 \
Signature · 1 line
2.17.1
#1 lore raw

PATCH v3 1/1 package/libnss: use correct CFLAGS and LDFLAGS

Hello Giulio,

Quoted 12 lines · On Mon, 29 Jul 2019 16:26:10 +0200 Giulio Benetti…

Actually libnss is configure with BUILD_OPT=1 that doesn't take into account TARGET_CFLAGS nor TARGET_LDFLAGS resulting in pre-chosen optimizations and debugging symbols.

Can't pass TARGET_CFLAGS through CFLAGS or similar otherwise other internal libnss one will be overwritten(i.e. -fPIC), so prefer to append TARGET_CFLAGS at the end of Linux.mk as well as TARGET_LDFLAGS according to internal libnss Makefile system's names. And obviously remove BUILD_OPT=1 from BUILD_VARS. This reveals hidden bugs when building with -Os due to uninitialized variables so add patches to fix those bugs.

Quoted 1 line

In fact, after inspecting more closely those uninitialized warnings from gcc, they are bogus. The variables are properly initialized, it's just gcc that goes nuts. So the proper fix is not to change the libnss code, and your patches are most likely going to be rejected upstream (with good reason).

So instead, I've dropped all of the patches, and simply passed NSS_ENABLE_WERROR=0 for both the target and host builds. This way -Werror is not passed, and those warnings become just warnings.

Applied with this change. Thanks!

Best regards,

Thomas

Signature · 3 lines
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
#2 lore raw

PATCH v3 1/1 package/libnss: use correct CFLAGS and LDFLAGS

Hello Thomas,

Il 01/08/2019 00:52, Thomas Petazzoni ha scritto:

Quoted 17 lines

Hello Giulio,

Quoted 12 lines · On Mon, 29 Jul 2019 16:26:10 +0200 Giulio Benetti…

Actually libnss is configure with BUILD_OPT=1 that doesn't take into account TARGET_CFLAGS nor TARGET_LDFLAGS resulting in pre-chosen optimizations and debugging symbols.

Can't pass TARGET_CFLAGS through CFLAGS or similar otherwise other internal libnss one will be overwritten(i.e. -fPIC), so prefer to append TARGET_CFLAGS at the end of Linux.mk as well as TARGET_LDFLAGS according to internal libnss Makefile system's names. And obviously remove BUILD_OPT=1 from BUILD_VARS. This reveals hidden bugs when building with -Os due to uninitialized variables so add patches to fix those bugs.

Quoted 5 lines
Quoted 1 line

In fact, after inspecting more closely those uninitialized warnings from gcc, they are bogus. The variables are properly initialized, it's just gcc that goes nuts.

This is what I thought in the beginning but after trying and retrying I've found that those warnings were correct! And indeed...

Quoted 3 lines

So the proper fix is not to change the libnss
code, and your patches are most likely going to be rejected upstream (with good reason).

...they have all been upstreamed:
https://hg.mozilla.org/projects/nss/log?rev=Giulio+Benetti

They all have been discussed here:
https://bugzilla.mozilla.org/buglist.cgi?emailcc1=1&list_id=14831966&emaillongdesc1=1&emailtype1=exact&emailassigned_to1=1&short_desc_type=allwordssubstr&email1=giulio.benetti%40micronovasrl.com&emailbug_mentor1=1&emailreporter1=1&short_desc=Remove%20-Wmaybe-uninitialized%20warning&resolution=FIXED&classification=Client%20Software&classification=Developer%20Infrastructure&classification=Components&classification=Server%20Software&classification=Other&query_format=advanced&emailqa_contact1=1&product=NSS

The first and more ambiguos warning was this:
https://bugzilla.mozilla.org/show_bug.cgi?id=1561548
It sent me nut, not gcc :)
But in the end it was correct.

Quoted 3 lines

So instead, I've dropped all of the patches, and simply passed NSS_ENABLE_WERROR=0 for both the target and host builds. This way -Werror is not passed, and those warnings become just warnings.

I'd say to keep as it is for the moment and when NSS_3_46 will be released(soon) it will contain all patches we will remove NSS_ENABLE_WERROR=0.

What do you think?

Signature · 10 lines
Giulio Benetti
CTO

MICRONOVA SRL
Sede: Via A. Niedda 3 - 35010 Vigonza (PD)
Tel. 049/8931563 - Fax 049/8931346
Cod.Fiscale - P.IVA 02663420285
Capitale Sociale ? 26.000 i.v.
Iscritta al Reg. Imprese di Padova N. 02663420285
Numero R.E.A. 258642
#3 lore raw

PATCH v3 1/1 package/libnss: use correct CFLAGS and LDFLAGS

Quoted 7 lines · On Thu, 1 Aug 2019 11:12:09 +0200 Giulio Benetti…
Quoted 3 lines

So the proper fix is not to change the libnss
code, and your patches are most likely going to be rejected upstream (with good reason).

...they have all been upstreamed:
https://hg.mozilla.org/projects/nss/log?rev=Giulio+Benetti

Quoted 7 lines

They all have been discussed here:
https://bugzilla.mozilla.org/buglist.cgi?emailcc1=1&list_id=14831966&emaillongdesc1=1&emailtype1=exact&emailassigned_to1=1&short_desc_type=allwordssubstr&email1=giulio.benetti%40micronovasrl.com&emailbug_mentor1=1&emailreporter1=1&short_desc=Remove%20-Wmaybe-uninitialized%20warning&resolution=FIXED&classification=Client%20Software&classification=Developer%20Infrastructure&classification=Components&classification=Server%20Software&classification=Other&query_format=advanced&emailqa_contact1=1&product=NSS

The first and more ambiguos warning was this:
https://bugzilla.mozilla.org/show_bug.cgi?id=1561548
It sent me nut, not gcc :)
But in the end it was correct.

I checked 3 of these (not all), and the three I checked were bogus, the variable was definitely initialized.

Quoted 5 lines

I'd say to keep as it is for the moment and when NSS_3_46 will be released(soon) it will contain all patches we will remove NSS_ENABLE_WERROR=0.

What do you think?

We generally disable Werror anyway in many packages, it's pretty common practice. It's common for new compiler versions to raise new warnings. Werror makes sense for upstream developers, a bit less so for downstream distributions/build-systems.

Best regards,

Thomas

Signature · 3 lines
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
#4 lore raw

PATCH v3 1/1 package/libnss: use correct CFLAGS and LDFLAGS

Il 01/08/2019 11:46, Thomas Petazzoni ha scritto:

Quoted 13 lines
Quoted 10 lines · On Thu, 1 Aug 2019 11:12:09 +0200 Giulio Benetti…
Quoted 3 lines

So the proper fix is not to change the libnss
code, and your patches are most likely going to be rejected upstream (with good reason).

...they have all been upstreamed:
https://hg.mozilla.org/projects/nss/log?rev=Giulio+Benetti

They all have been discussed here:
https://bugzilla.mozilla.org/buglist.cgi?emailcc1=1&list_id=14831966&emaillongdesc1=1&emailtype1=exact&emailassigned_to1=1&short_desc_type=allwordssubstr&email1=giulio.benetti%40micronovasrl.com&emailbug_mentor1=1&emailreporter1=1&short_desc=Remove%20-Wmaybe-uninitialized%20warning&resolution=FIXED&classification=Client%20Software&classification=Developer%20Infrastructure&classification=Components&classification=Server%20Software&classification=Other&query_format=advanced&emailqa_contact1=1&product=NSS

Quoted 7 lines
Quoted 4 lines

The first and more ambiguos warning was this:
https://bugzilla.mozilla.org/show_bug.cgi?id=1561548
It sent me nut, not gcc :)
But in the end it was correct.

I checked 3 of these (not all), and the three I checked were bogus, the variable was definitely initialized.

You're right, I've been misleaded, indeed my commit log doesn't say that they are real errors.

Quoted 10 lines
Quoted 5 lines

I'd say to keep as it is for the moment and when NSS_3_46 will be released(soon) it will contain all patches we will remove NSS_ENABLE_WERROR=0.

What do you think?

We generally disable Werror anyway in many packages, it's pretty common practice. It's common for new compiler versions to raise new warnings. Werror makes sense for upstream developers, a bit less so for downstream distributions/build-systems.

Ok, thank you!

Best regards

Signature · 10 lines
Giulio Benetti
CTO

MICRONOVA SRL
Sede: Via A. Niedda 3 - 35010 Vigonza (PD)
Tel. 049/8931563 - Fax 049/8931346
Cod.Fiscale - P.IVA 02663420285
Capitale Sociale ? 26.000 i.v.
Iscritta al Reg. Imprese di Padova N. 02663420285
Numero R.E.A. 258642
#5 lore raw

PATCH v3 1/1 package/libnss: use correct CFLAGS and LDFLAGS

In reply to Giulio Benetti
Quoted 14 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Giulio" == Giulio Benetti <[email protected]> writes:

Actually libnss is configure with BUILD_OPT=1 that doesn't take into account TARGET_CFLAGS nor TARGET_LDFLAGS resulting in pre-chosen optimizations and debugging symbols.

Can't pass TARGET_CFLAGS through CFLAGS or similar otherwise other internal libnss one will be overwritten(i.e. -fPIC), so prefer to append TARGET_CFLAGS at the end of Linux.mk as well as TARGET_LDFLAGS according to internal libnss Makefile system's names. And obviously remove BUILD_OPT=1 from BUILD_VARS. This reveals hidden bugs when building with -Os due to uninitialized variables so add patches to fix those bugs.

Quoted 1 line

Committed to 2019.02.x and 2019.05.x, thanks (after the changes done by Thomas to not patch).

Signature · 1 line
Bye, Peter Korsgaard