librelp: upgrade from 1.2.12 to 1.2.14
PATCH 1/2 librelp: upgrade from 1.2.12 to 1.2.14
---
.../recipes-extended/rsyslog/{librelp_1.2.12.bb => librelp_1.2.14.bb} | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
rename meta-oe/recipes-extended/rsyslog/{librelp_1.2.12.bb => librelp_1.2.14.bb} (86%)meta-oe/recipes-extended/rsyslog/librelp_1.2.14.bb renamed from meta-oe/recipes-extended/rsyslog/librelp_1.2.12.bb +1 -1
@@ -8,7 +8,7 @@ DEPENDS = "gmp nettle libidn zlib gnutls" SRC_URI = "git://github.com/rsyslog/librelp.git;protocol=https" -SRCREV = "02c3be4f5c39fec59d05cd8b75b08dbba04098ad" +SRCREV = "fc512e337bfc7c92770246dbff5f482b879498b9" S = "${WORKDIR}/git"
Signature · 1 line
2.7.4
PATCH 2/2 rsyslog: update from 8.22 to 8.29
Drop two patches:
CVE-2017-12588.patch
0001-core-bugfix-segfault-after-configuration-errors.patchsince they are included in 8.29.
Update the testbench configuration flags.
---
...ugfix-segfault-after-configuration-errors.patch | 90 ----------------------
.../rsyslog/rsyslog/CVE-2017-12588.patch | 40 ----------
.../{rsyslog_8.22.0.bb => rsyslog_8.29.0.bb} | 8 +-
3 files changed, 3 insertions(+), 135 deletions(-)
delete mode 100644 meta-oe/recipes-extended/rsyslog/rsyslog/0001-core-bugfix-segfault-after-configuration-errors.patch
delete mode 100644 meta-oe/recipes-extended/rsyslog/rsyslog/CVE-2017-12588.patch
rename meta-oe/recipes-extended/rsyslog/{rsyslog_8.22.0.bb => rsyslog_8.29.0.bb} (94%)meta-oe/recipes-extended/rsyslog/rsyslog/0001-core-bugfix-segfault-after-configuration-errors.patch deleted +0 -90
@@ -1,90 +0,0 @@ -From 6d258339802cb9f13d8a4a157a4b74eccb902d8f Mon Sep 17 00:00:00 2001 -From: Rainer Gerhards <[email protected]> -Date: Mon, 17 Jul 2017 15:36:32 +0200 -Subject: [PATCH] core bugfix: segfault after configuration errors - -rsyslog will segfault on startup if -a) the local machine's hostname is set to a non-FQDN name -b) the getaddrinfo() system call fails -This scenario is higly unlikely, but may exist especially with -provisioned VMs which may not properly be able to do name queries -on startup (seen for example on AWS). - -This patch fixes the situation and also provides more robustness -for very early startup error messages when some of the error-reporting -subsystem is not yet properly initialized. Note that under these -circumstances, errors may only show up on stderr. - -Upstream status: Backport - -closes https://github.com/rsyslog/rsyslog/issues/1573 - -Signed-off-by: Ovidiu Panait <[email protected]> ---- - runtime/prop.c | 6 ++++++ - tools/rsyslogd.c | 17 +++++++++-------- - 2 files changed, 15 insertions(+), 8 deletions(-) - -diff --git a/runtime/prop.c b/runtime/prop.c -index e5b4693..cb93285 100644 ---- a/runtime/prop.c -+++ b/runtime/prop.c -@@ -133,7 +133,13 @@ propConstructFinalize(prop_t __attribute__((unused)) *pThis) - */ - static rsRetVal AddRef(prop_t *pThis) - { -+ if(pThis == NULL) { -+ DBGPRINTF("prop/AddRef is passed a NULL ptr - ignoring it " -+ "- further problems may occur\n"); -+ FINALIZE; -+ } - ATOMIC_INC(&pThis->iRefCount, &pThis->mutRefCount); -+finalize_it: - return RS_RET_OK; - } - -diff --git a/tools/rsyslogd.c b/tools/rsyslogd.c -index 759d293..6aa1487 100644 ---- a/tools/rsyslogd.c -+++ b/tools/rsyslogd.c -@@ -808,9 +808,11 @@ logmsgInternal(int iErr, const syslog_pri_t pri, const uchar *const msg, int fla - * permits us to process unmodified config files which otherwise contain a - * supressor statement. - */ -- if(((Debug == DEBUG_FULL || !doFork) && ourConf->globals.bErrMsgToStderr) || iConfigVerify) { -+ int emit_to_stderr = (ourConf == NULL) ? 1 : ourConf->globals.bErrMsgToStderr; -+ if(((Debug == DEBUG_FULL || !doFork) && emit_to_stderr) || iConfigVerify) { - if(pri2sev(pri) == LOG_ERR) -- fprintf(stderr, "rsyslogd: %s\n", (bufModMsg == NULL) ? (char*)msg : bufModMsg); -+ fprintf(stderr, "rsyslogd: %s\n", -+ (bufModMsg == NULL) ? (char*)msg : bufModMsg); - } - - finalize_it: -@@ -1115,18 +1117,17 @@ initAll(int argc, char **argv) - - /* doing some core initializations */ - -- /* get our host and domain names - we need to do this early as we may emit -- * error log messages, which need the correct hostname. -- rgerhards, 2008-04-04 -- */ -- queryLocalHostname(); -- -- /* initialize the objects */ - if((iRet = modInitIminternal()) != RS_RET_OK) { - fprintf(stderr, "fatal error: could not initialize errbuf object (error code %d).\n", - iRet); - exit(1); /* "good" exit, leaving at init for fatal error */ - } - -+ /* get our host and domain names - we need to do this early as we may emit -+ * error log messages, which need the correct hostname. -- rgerhards, 2008-04-04 -+ * But we need to have imInternal up first! -+ */ -+ queryLocalHostname(); - - /* END core initializations - we now come back to carrying out command line options*/ - --- -2.10.2 -
meta-oe/recipes-extended/rsyslog/rsyslog/CVE-2017-12588.patch deleted +0 -40
@@ -1,40 +0,0 @@ -From 6bc4aa975a83abed43d734299ce76cd9e1a14aec Mon Sep 17 00:00:00 2001 -From: Thomas Deutschmann <[email protected]> -Date: Wed, 17 May 2017 23:05:24 +0200 -Subject: [PATCH] imzmq3: Fix building with -Werror=format-security - -Reference: https://nvd.nist.gov/vuln/detail/CVE-2017-12588 - -CVE: 2017-12588 - -Upstream-Status: Backport -Signed-off-by: Chen Qi <[email protected]> ---- - contrib/imzmq3/imzmq3.c | 4 ++-- - 1 file changed, 2 insertions(+), 2 deletions(-) - -diff --git a/contrib/imzmq3/imzmq3.c b/contrib/imzmq3/imzmq3.c -index 9ca17871..d32dcbc2 100644 ---- a/contrib/imzmq3/imzmq3.c -+++ b/contrib/imzmq3/imzmq3.c -@@ -403,7 +403,7 @@ static rsRetVal createSocket(instanceConf_t* info, void** sock) { - - /* Do the bind/connect... */ - if (info->action==ACTION_CONNECT) { -- rv = zsocket_connect(*sock, info->description); -+ rv = zsocket_connect(*sock, "%s", info->description); - if (rv == -1) { - errmsg.LogError(0, - RS_RET_INVALID_PARAMS, -@@ -413,7 +413,7 @@ static rsRetVal createSocket(instanceConf_t* info, void** sock) { - } - DBGPRINTF("imzmq3: connect for %s successful\n",info->description); - } else { -- rv = zsocket_bind(*sock, info->description); -+ rv = zsocket_bind(*sock, "%s", info->description); - if (rv == -1) { - errmsg.LogError(0, - RS_RET_INVALID_PARAMS, --- -2.13.0 -
meta-oe/recipes-extended/rsyslog/rsyslog_8.29.0.bb renamed from meta-oe/recipes-extended/rsyslog/rsyslog_8.22.0.bb +3 -5
@@ -24,8 +24,6 @@ SRC_URI = "http://www.rsyslog.com/download/files/download/rsyslog/${BPN}-${PV}.t file://use-pkgconfig-to-check-libgcrypt.patch \ file://run-ptest \ file://rsyslog-fix-ptest-not-finish.patch \ - file://CVE-2017-12588.patch \ - file://0001-core-bugfix-segfault-after-configuration-errors.patch \ " SRC_URI_append_libc-musl = " \
@@ -33,8 +31,8 @@ SRC_URI_append_libc-musl = " \ file://0001-Include-sys-time-h.patch \ " -SRC_URI[md5sum] = "ad0f25f429aa2daa326732950a5eeb6c" -SRC_URI[sha256sum] = "06e2884181333dccecceaca82827ae24ca7a258b4fbf7b1e07a80d4caae640ca" +SRC_URI[md5sum] = "3805617f65a4b4bea34606487a5255a0" +SRC_URI[sha256sum] = "220ba30b5afb0f3ddb328613fea7aa3966b01e4d0c52d6de9ab27b0858f19738" inherit autotools pkgconfig systemd update-rc.d ptest
@@ -57,7 +55,7 @@ PACKAGECONFIG[klog] = "--enable-klog,--disable-klog,," PACKAGECONFIG[regexp] = "--enable-regexp,--disable-regexp,," PACKAGECONFIG[uuid] = "--enable-uuid,--disable-uuid,util-linux," PACKAGECONFIG[libgcrypt] = "--enable-libgcrypt,--disable-libgcrypt,libgcrypt," -PACKAGECONFIG[testbench] = "--enable-testbench,--disable-testbench,," +PACKAGECONFIG[testbench] = "--enable-testbench --enable-omstdout,--disable-testbench --disable-omstdout,," # default no in configure PACKAGECONFIG[debug] = "--enable-debug,--disable-debug,,"
Signature · 1 line
2.7.4
Re: [PATCH 2/2] rsyslog: update from 8.22 to 8.29
Quoted 8 lines · Randy MacLeod
Drop two patches:
CVE-2017-12588.patch 0001-core-bugfix-segfault-after-configuration-errors.patchsince they are included in 8.29.
Update the testbench configuration flags.
Signed-off-by Randy MacLeod <[email protected]>
...
rsyslog-8.29 builds for all arches and for qemux86-64/arm64 using musl. basic testing with systemd works and I can restart rsyslog.
On qemux86-64 ptest runs but encounters 5 failures:
============================================================================
Testsuite summary for rsyslog 8.29.0
============================================================================
# TOTAL: 265
# PASS: 246
# SKIP: 14
# XFAIL: 0
# FAIL: 5
# XPASS: 0
# ERROR: 0
# systemd-related problems I think. FAIL: omfile-read-only-errmsg.sh FAIL: omfile-read-only.sh ... # these two are related to the testbench user mgmt FAIL: privdropuser.sh FAIL: privdropuserid.sh ... # this test assumes that there's a 'service' command FAIL: imklog_permitnonkernelfacility_root.sh
I don't see anything that would prevent us from taking the upgrade into master.
I might fix these ptests failures or just skip them but not for a while.
Signature · 2 lines
# Randy MacLeod. WR Linux # Wind River an Intel Company
Re: [PATCH 2/2] rsyslog: update from 8.22 to 8.29
Quoted 42 lines · Randy MacLeod
Quoted 8 lines · Randy MacLeod
Drop two patches:
CVE-2017-12588.patch 0001-core-bugfix-segfault-after-configuration-errors.patchsince they are included in 8.29.
Update the testbench configuration flags.
Signed-off-by Randy MacLeod <[email protected]>...
rsyslog-8.29 builds for all arches and for qemux86-64/arm64 using musl. basic testing with systemd works and I can restart rsyslog.On qemux86-64 ptest runs but encounters 5 failures:
============================================================================
Testsuite summary for rsyslog 8.29.0
============================================================================# TOTAL: 265
# PASS: 246
# SKIP: 14
# XFAIL: 0
# FAIL: 5
# XPASS: 0
# ERROR: 0# systemd-related problems I think. FAIL: omfile-read-only-errmsg.sh FAIL: omfile-read-only.sh ... # these two are related to the testbench user mgmt FAIL: privdropuser.sh FAIL: privdropuserid.sh ... # this test assumes that there's a 'service' command FAIL: imklog_permitnonkernelfacility_root.sh
Quoted 5 lines
I don't see anything that would prevent us from taking the upgrade into master.
I might fix these ptests failures or just skip them but not for a while.
I switched to sysvinit and a build that has the 'service' wrapper (our wrlinux-image-glibc-std) and:
============================================================================
Testsuite summary for rsyslog 8.29.0
============================================================================
# TOTAL: 265
# PASS: 247
# SKIP: 14
# XFAIL: 0
# FAIL: 4
# XPASS: 0
# ERROR: 0
FAIL: omfile-read-only-errmsg.sh
FAIL: omfile-read-only.sh
...
FAIL: privdropuser.sh
FAIL: privdropuserid.sh
so much for most of that theory explaining the first to failures but the last one goes away as expected.
Signature · 2 lines
# Randy MacLeod. WR Linux # Wind River an Intel Company