Summary: | net-mgmt/pmacct: Update to 1.7.7 | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Product: | Ports & Packages | Reporter: | Robert Clausecker <fuz> | ||||||||
Component: | Individual Port(s) | Assignee: | Kurt Jaeger <pi> | ||||||||
Status: | Closed FIXED | ||||||||||
Severity: | Affects Only Me | CC: | pi | ||||||||
Priority: | --- | Flags: | pi:
maintainer-feedback+
fuz: merge-quarterly? |
||||||||
Version: | Latest | ||||||||||
Hardware: | Any | ||||||||||
OS: | Any | ||||||||||
URL: | http://www.pmacct.net/ChangeLog-1.7.7 | ||||||||||
Attachments: |
|
Description
Robert Clausecker
2021-12-03 23:06:12 UTC
Comment on attachment 229865 [details]
net-mgmt/pmacct: update to 1.7.7
maintainer timeout
maintainer timeout Thanks for the patch. It fails to build on i386, because of this: sql_common_m.c:262:48: error: format specifies type 'unsigned long' but the argument has type 'time_t' (aka 'int') [-Werror,-Wformat] snprintf(tmpptr, envbuf.end-tmpptr, "%lu", idata->elap_time); ~~~ ^~~~~~~~~~~~~~~~ %d sql_common_m.c:295:48: error: format specifies type 'unsigned long' but the argument has type 'time_t' (aka 'int') [-Werror,-Wformat] snprintf(tmpptr, envbuf.end-tmpptr, "%lu", idata->basetime); ~~~ ^~~~~~~~~~~~~~~ %d sql_common_m.c:306:48: error: format specifies type 'unsigned long' but the argument has type 'time_t' (aka 'int') [-Werror,-Wformat] snprintf(tmpptr, envbuf.end-tmpptr, "%lu", idata->timeslot); ~~~ ^~~~~~~~~~~~~~~ %d 3 errors generated. It's fine otherwise. I try to find a way to find a patch that is applied on i386 only. (In reply to Kurt Jaeger from comment #3) Hi Kurt, No need to that. I'll update the patch such that it does not fail on i386 either. It should be as simple as casting idata->elap_time, idata->basetime, and idata->timeslot to (long). Created attachment 231364 [details]
extra-patch-i386
Together with some change to the ports Makefile, this patch works.
Looks good! Committed, thanks! A commit in branch main references this bug: URL: https://cgit.FreeBSD.org/ports/commit/?id=c2da42cc80a9b3f707f6186c1e8c4ecbf1efb0f1 commit c2da42cc80a9b3f707f6186c1e8c4ecbf1efb0f1 Author: Robert Clausecker <fuz@fuz.su> AuthorDate: 2022-01-26 20:03:41 +0000 Commit: Kurt Jaeger <pi@FreeBSD.org> CommitDate: 2022-01-26 20:03:41 +0000 net-mgmt/pmacct: update 1.7.5 -> 1.7.7 - remove -Werror as per policy - touch up patches - remove PLABEL option for feature removed in 1.7.6 PR: 260194 Changes: http://www.pmacct.net/ChangeLog-1.7.7 net-mgmt/pmacct/Makefile | 20 ++++++------- net-mgmt/pmacct/distinfo | 6 ++-- net-mgmt/pmacct/files/extra-patch-i386 (new) | 29 ++++++++++++++++++ net-mgmt/pmacct/files/patch-Makefile.in | 4 +-- net-mgmt/pmacct/files/patch-src_nfacctd.c (gone) | 11 ------- .../files/patch-src_nfprobe__plugin_netflow9.c | 4 +-- net-mgmt/pmacct/files/patch-src_pgsql__plugin.c | 4 +-- net-mgmt/pmacct/files/patch-src_pmacct.c | 8 ++--- net-mgmt/pmacct/files/patch-src_pmacctd.c (gone) | 11 ------- net-mgmt/pmacct/files/patch-src_pmbgpd.c (gone) | 11 ------- net-mgmt/pmacct/files/patch-src_pmbmpd.c (gone) | 11 ------- .../pmacct/files/patch-src_pmtelemetryd.c (gone) | 11 ------- net-mgmt/pmacct/files/patch-src_sfacctd.c (gone) | 11 ------- .../pmacct/files/patch-src_sql__common.c (gone) | 34 ---------------------- net-mgmt/pmacct/files/patch-src_sqlite3__plugin.c | 4 +-- net-mgmt/pmacct/files/patch-src_util.c | 13 ++------- net-mgmt/pmacct/pkg-plist | 2 ++ 17 files changed, 58 insertions(+), 136 deletions(-) The port still fails to build on armv7 because you removed the part of my patch that removes -Werror. Why did you do that? -Werror is against policy as it leads to spurious build failures due to often benign things. Now I have to produce yet another patch. This sucks! Reopening as the port still fails to build on armv7 due to a botched patch application. Created attachment 234286 [details]
net-mgmt/pmacct: fix build on armv7
Here is another patch which has the parts you decided not to commit in the first round. These are actually important. Spurious warnings tend to cause random build failures so -Werror must be patched out. I don't want to have to come and submit another patch.
The other part of the patch fixes a wrong type in a log call. This causes a problem when the bgp session is more than 2^31 - 1 seconds out of date in which case it'll not show the correct amount of seconds by which it is out of date. Terrifying issue, isn't it? Good thing -Werror caught this one!
(In reply to Robert Clausecker from comment #9) Yes, I messed up the commit by missing out the two patch files. I'll work to get your patch in. Committed fix for armv7. A commit in branch main references this bug: URL: https://cgit.FreeBSD.org/ports/commit/?id=16f00133d92bda820f660e0c51575865d7a259b7 commit 16f00133d92bda820f660e0c51575865d7a259b7 Author: Robert Clausecker <fuz@fuz.su> AuthorDate: 2022-06-12 19:53:01 +0000 Commit: Kurt Jaeger <pi@FreeBSD.org> CommitDate: 2022-06-12 19:53:01 +0000 net-mgmt/pmacct: fix build on armv7 PR: 260194 Author: Robert Clausecker <fuz@fuz.su> net-mgmt/pmacct/Makefile | 2 +- net-mgmt/pmacct/files/patch-configure.ac (new) | 20 ++++++++++++++++++++ net-mgmt/pmacct/files/patch-src_bgp_bgp.c (new) | 14 ++++++++++++++ .../files/patch-src_external__libs_Makefile.am (new) | 16 ++++++++++++++++ 4 files changed, 51 insertions(+), 1 deletion(-) And you've still not patched out -Werror. See you next update I guess. :-( |