Add Earthly-based builds #2
Conversation
Currently builds only ubuntu 20.04 packages
Threw in amd64/arm64 multi-platform builds
alistairking
force-pushed
the
a6r--earthly
branch
from
06c8acf
to
cd002ec
Compare
|
Very nice! For testing the built package, see "Test Steps" in our building_frr doc:
|
|
Thanks again for this PR @alistairking! I'm happy to be getting around to testing earthly on my Windows Desktop on my home network -- first time ran through I hit the following error: As a workaround, I'm tempted to try Earthly's "GitHub Actions integration paired with the upload-artifact GitHub Action. I'll come back to dig into this later! Earthly build failure +deps-one | E: Failed to fetch http://archive.ubuntu.com/ubuntu/pool/main/p/po-debconf/po-debconf_1.0.21_all.deb Undetermined Error [IP: 185.125.190.36 80]
+deps-one | E: Unable to fetch some archives, maybe run apt-get update or try with --fix-missing?
+deps-one | ERROR Earthfile line 79:4
+deps-one | The command
+deps-one | RUN DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends autoconf automake libtool make build-essential devscripts equivs libreadline-dev texinfo pkg-config libpam0g-dev libjson-c-dev bison flex libc-ares-dev python3-dev python3-sphinx install-info libsnmp-dev perl libcap-dev libelf-dev libunwind-dev wget
+deps-one | did not complete successfully. Exit code 100
+pkg-one | WARN: Canceled
================ ❌ FAILURE [2. Build 🔧] ================ |
|
Any chance you're behind some kind of proxy? |
Not intentionally behind a proxy on my home network... 🤔 I wonder if 'UO Split VPN' solution that I have running is effecting this build. I'll try to kick off another build when I'm not using that VPN later today. |
|
Running again from my home network without VPN -- running into a similar network issue but with a different package this time: Now I'm even more tempted to try Earthly's "GitHub Actions integration paired with the upload-artifact GitHub Action... I'm going to want to look into Limits for GitHub Actions probably to make sure I don't drain all of Route Views' freebie minutes 😅 |
Fixing the crash: > #0 0x0000560aa80f8e30 in lspdb_const_find (h=<error reading variable: Cannot access memory at address 0x7fff5e95efe8>, item=<error reading variable: Cannot access memory at address 0x7fff5e95efe0>) at ./isisd/isis_lsp.h:64 > #1 0x0000560aa80f8e9d in lspdb_find (h=0x560aaa1ed3b8, item=0x7fff5e95f050) at ./isisd/isis_lsp.h:64 > #2 0x0000560aa80f92f9 in lsp_search (head=0x560aaa1ed3b8, id=0x7fff5e95f200 "") at isisd/isis_lsp.c:100 > FRRouting#3 0x0000560aa8113d69 in spf_adj_list_parse_tlv (spftree=0x560aaa1f09d0, adj_list=0x560aaa214480, id=0x560aad331a78 "", desig_is_id=0x0, pseudo_metric=0, metric=3, oldmetric=false, subtlvs=0x0) at isisd/isis_spf.c:1330 > FRRouting#4 0x0000560aa811419d in spf_adj_list_parse_lsp (spftree=0x560aaa1f09d0, adj_list=0x560aaa214480, lsp=0x560aaa1f4e50, pseudo_nodeid=0x0, pseudo_metric=0) at isisd/isis_spf.c:1429 > FRRouting#5 0x0000560aa81141fe in spf_adj_list_parse_lsp (spftree=0x560aaa1f09d0, adj_list=0x560aaa214480, lsp=0x560aaa1ff8e0, pseudo_nodeid=0x0, pseudo_metric=0) at isisd/isis_spf.c:1442 > FRRouting#6 0x0000560aa81141fe in spf_adj_list_parse_lsp (spftree=0x560aaa1f09d0, adj_list=0x560aaa214480, lsp=0x560aaa1f4e50, pseudo_nodeid=0x0, pseudo_metric=0) at isisd/isis_spf.c:1442 > (...) > #65507 0x0000560aa81141fe in spf_adj_list_parse_lsp (spftree=0x560aaa1f09d0, adj_list=0x560aaa214480, lsp=0x560aaa1ff8e0, pseudo_nodeid=0x0, pseudo_metric=0) at isisd/isis_spf.c:1442 > #65508 0x0000560aa81141fe in spf_adj_list_parse_lsp (spftree=0x560aaa1f09d0, adj_list=0x560aaa214480, lsp=0x560aaa1f4e50, pseudo_nodeid=0x0, pseudo_metric=0) at isisd/isis_spf.c:1442 > #65509 0x0000560aa81141fe in spf_adj_list_parse_lsp (spftree=0x560aaa1f09d0, adj_list=0x560aaa214480, lsp=0x560aaa1ff8e0, pseudo_nodeid=0x0, pseudo_metric=0) at isisd/isis_spf.c:1442 > #65510 0x0000560aa81141fe in spf_adj_list_parse_lsp (spftree=0x560aaa1f09d0, adj_list=0x560aaa214480, lsp=0x560aaa1f4e50, pseudo_nodeid=0x0, pseudo_metric=0) at isisd/isis_spf.c:1442 > #65511 0x0000560aa8114313 in isis_spf_build_adj_list (spftree=0x560aaa1f09d0, lsp=0x560aaa1f4e50) at isisd/isis_spf.c:1455 > #65512 0x0000560aa8114f09 in isis_run_spf (spftree=0x560aaa1f09d0) at isisd/isis_spf.c:1775 > #65513 0x0000560aa8115057 in isis_run_spf_with_protection (area=0x560aaa1ed3b0, spftree=0x560aaa1f09d0) at isisd/isis_spf.c:1801 > #65514 0x0000560aa8115311 in isis_run_spf_cb (thread=0x7fff5f15e5a0) at isisd/isis_spf.c:1859 > #65515 0x00007f90bac66dcc in thread_call (thread=0x7fff5f15e5a0) at lib/thread.c:2002 > #65516 0x00007f90bac013ee in frr_run (master=0x560aa9f5cb40) at lib/libfrr.c:1196 > #65517 0x0000560aa80e7da2 in main (argc=2, argv=0x7fff5f15e7b8, envp=0x7fff5f15e7d0) at isisd/isis_main.c:273 The fix is similar to the crash fix included in d9884a7 ("isisd: Prepare IS-IS for Link State support"). The fix was: > diff --git a/isisd/isis_lsp.c b/isisd/isis_lsp.c > index 94353a5bc8..92d329f035 100644 > --- a/isisd/isis_lsp.c > +++ b/isisd/isis_lsp.c > @@ -2166,7 +2178,7 @@ int isis_lsp_iterate_ip_reach(struct isis_lsp *lsp, int family, uint16_t mtid, > if (lsp->hdr.seqno == 0 || lsp->hdr.rem_lifetime == 0) > return LSP_ITER_CONTINUE; > > - /* Parse main LSP. */ > + /* Parse LSP */ > if (lsp->tlvs) { > if (!fabricd && !pseudo_lsp && family == AF_INET > && mtid == ISIS_MT_IPV4_UNICAST) { > @@ -2236,13 +2248,17 @@ int isis_lsp_iterate_ip_reach(struct isis_lsp *lsp, int family, uint16_t mtid, > } > } > > - /* Parse LSP fragments. */ > - for (ALL_LIST_ELEMENTS_RO(lsp->lspu.frags, node, frag)) { > - if (!frag->tlvs) > - continue; > + /* Parse LSP fragments if it is not a fragment itself */ > + if (!LSP_FRAGMENT(lsp->hdr.lsp_id)) > + for (ALL_LIST_ELEMENTS_RO(lsp->lspu.frags, node, frag)) { > + if (!frag->tlvs) > + continue; > > - isis_lsp_iterate_ip_reach(frag, family, mtid, cb, arg); > - } > + if (isis_lsp_iterate_ip_reach(frag, family, mtid, cb, > + arg) > + == LSP_ITER_STOP) > + return LSP_ITER_STOP; > + } > > return LSP_ITER_CONTINUE; > } Fixes: 7b36d36 ("isisd: make the SPF code more modular") Fixes: 5e56a50 ("isisd: fix infinite loop when parsing LSPs") Signed-off-by: Louis Scalbert <[email protected]> (cherry picked from commit 8c8a5a0)
|
@alistairking, I've been working on a GitHub Action workflow for "Build FRR" in the However, the I wonder if you have thoughts on why I can't find packages from the Earthly build? (Maybe I do just need the "--push" flag, like they suggest in their template GitHub Action -- I'll read more and give that a try). |
|
You might want to take a look at: https://github.com/alistairking/scamper/blob/master/.github/workflows/build.yml |
Currently builds for Ubuntu 20.04, 22.04, Centos 7/8.
rvbranch eventually)earthly +all./dist/Results: