ci: failure in interface_usdt

fanquake commented at 8:55 am on April 22, 2025: member

https://github.com/bitcoin/bitcoin/actions/runs/14579386142/job/40892496346#step:6:5102:

 0 122/318 - interface_usdt_coinselection.py failed, Duration: 16 s
 1
 2stdout:
 32025-04-21T19:37:36.056000Z TestFramework (INFO): PRNG seed is: 4088013170701346751
 42025-04-21T19:37:36.066000Z TestFramework (INFO): Initializing test directory /home/runner/work/_temp/ci/scratch/test_runner/test_runner_₿_🏃_20250421_192058/interface_usdt_coinselection_192
 52025-04-21T19:37:38.378000Z TestFramework (INFO): hook into the coin_selection tracepoints
 62025-04-21T19:37:48.532000Z TestFramework (INFO): Prepare wallets
 72025-04-21T19:37:50.021000Z TestFramework (INFO): Sending a transaction should result in all tracepoints
 82025-04-21T19:37:50.114000Z TestFramework (INFO): Failing to fund results in 1 tracepoint
 92025-04-21T19:37:50.146000Z TestFramework (INFO): Explicitly enabling APS results in 2 tracepoints
102025-04-21T19:37:50.238000Z TestFramework (INFO): Change position is -1 if no change is created with APS when APS was initially not used
112025-04-21T19:37:50.523000Z TestFramework (INFO): Change position is -1 if no change is created normally and APS is not used
122025-04-21T19:37:50.772000Z TestFramework (INFO): Stopping nodes
132025-04-21T19:37:51.192000Z TestFramework (INFO): Cleaning up /home/runner/work/_temp/ci/scratch/test_runner/test_runner_₿_🏃_20250421_192058/interface_usdt_coinselection_192 on exit
142025-04-21T19:37:51.192000Z TestFramework (INFO): Tests successful
15
16
17stderr:
18In file included from /virtual/main.c:1:
19In file included from include/uapi/linux/ptrace.h:183:
20In file included from arch/x86/include/asm/ptrace.h:175:
21In file included from arch/x86/include/asm/paravirt_types.h:12:
22In file included from arch/x86/include/asm/nospec-branch.h:15:
23arch/x86/include/asm/current.h:47:10: warning: multiple identical address spaces specified for type [-Wduplicate-decl-specifier]
24   47 |                 return this_cpu_read_const(const_pcpu_hot.current_task);
25      |                        ^
26arch/x86/include/asm/percpu.h:577:36: note: expanded from macro 'this_cpu_read_const'
27  577 | #define this_cpu_read_const(pcp)                        __raw_cpu_read_const(pcp)
28      |                                                         ^
29arch/x86/include/asm/percpu.h:163:35: note: expanded from macro '__raw_cpu_read_const'
30  163 | #define __raw_cpu_read_const(pcp)       __raw_cpu_read(, , pcp)
31      |                                         ^
32arch/x86/include/asm/percpu.h:155:30: note: expanded from macro '__raw_cpu_read'
33  155 |         *(qual __my_cpu_type(pcp) *)__my_cpu_ptr(&(pcp));               \
34      |                                     ^
35note: (skipping 1 expansions in backtrace; use -fmacro-backtrace-limit=0 to see all)
36arch/x86/include/asm/percpu.h:94:40: note: expanded from macro '__my_cpu_type'
37   94 | #define __my_cpu_type(var)      typeof(var) __percpu_seg_override
38      |                                             ^
39arch/x86/include/asm/percpu.h:45:32: note: expanded from macro '__percpu_seg_override'
40   45 | # define __percpu_seg_override  __seg_gs
41      |                                 ^
42<built-in>:358:33: note: expanded from macro '__seg_gs'
43  358 | #define __seg_gs __attribute__((address_space(256)))
44      |                                 ^
45In file included from /virtual/main.c:1:
46In file included from include/uapi/linux/ptrace.h:183:
47In file included from arch/x86/include/asm/ptrace.h:175:
48In file included from arch/x86/include/asm/paravirt_types.h:12:
49In file included from arch/x86/include/asm/nospec-branch.h:15:
50arch/x86/include/asm/current.h:47:10: warning: multiple identical address spaces specified for type [-Wduplicate-decl-specifier]
51arch/x86/include/asm/percpu.h:577:36: note: expanded from macro 'this_cpu_read_const'
52  577 | #define this_cpu_read_const(pcp)                        __raw_cpu_read_const(pcp)
53      |                                                         ^
54arch/x86/include/asm/percpu.h:163:35: note: expanded from macro '__raw_cpu_read_const'
55  163 | #define __raw_cpu_read_const(pcp)       __raw_cpu_read(, , pcp)
56      |                                         ^
57arch/x86/include/asm/percpu.h:155:9: note: expanded from macro '__raw_cpu_read'
58  155 |         *(qual __my_cpu_type(pcp) *)__my_cpu_ptr(&(pcp));               \
59      |                ^
60arch/x86/include/asm/percpu.h:94:40: note: expanded from macro '__my_cpu_type'
61   94 | #define __my_cpu_type(var)      typeof(var) __percpu_seg_override
62      |                                             ^
63arch/x86/include/asm/percpu.h:45:32: note: expanded from macro '__percpu_seg_override'
64   45 | # define __percpu_seg_override  __seg_gs
65      |                                 ^
66<built-in>:358:33: note: expanded from macro '__seg_gs'
67  358 | #define __seg_gs __attribute__((address_space(256)))
68      |                                 ^
692 warnings generated.

fanquake commented at 8:55 am on April 22, 2025: member

cc @0xB10C

maflcko commented at 9:02 am on April 22, 2025: member

Probably an erroneous kernel that was pushed to GHA by microsoft?

maflcko added the label CI failed on Apr 22, 2025

maflcko commented at 6:18 pm on April 22, 2025: member

Passing one: https://github.com/bitcoin/bitcoin/actions/runs/14580801695/job/40919798022#step:6:2537

System info: Linux 6.8.0-1021-azure

Failing one: https://github.com/bitcoin/bitcoin/actions/runs/14580801695/job/40896887951#step:6:2538

System info: Linux 6.11.0-1012-azure

ref: https://github.com/actions/runner-images/pull/12065

0xB10C commented at 10:08 am on April 23, 2025: contributor

Seems like a one-off event for now and not much that can be done about it?

hodlinator commented at 12:06 pm on April 23, 2025: contributor

Happened again here: https://github.com/hodlinator/bitcoin/actions/runs/14614533923/job/40999660639#step:6:5096

maflcko commented at 12:28 pm on April 23, 2025: member

Seems like a one-off event for now and not much that can be done about it?

No, this happens consistently with the new kernel.

I suspect anyone can reproduce this with a new kernel locally as well.

My suggestion would be to fix this or work around it in some way temporarily, until the kernel takes the patch (one was submitted iirc).

If it isn’t possible to fix or work around, the test should be disabled temporarily:

 0    ci: Temporarily disable failing bpf checks
 1
 2diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
 3index 9d8fdb4a16..13f86dd191 100644
 4--- a/.github/workflows/ci.yml
 5+++ b/.github/workflows/ci.yml
 6@@ -405,11 +405,6 @@ jobs:
 7           key: ${{ github.job }}-ccache-${{ github.run_id }}
 8           restore-keys: ${{ github.job }}-ccache-
 9 
10-      - name: Enable bpfcc script
11-        # In the image build step, no external environment variables are available,
12-        # so any settings will need to be written to the settings env file:
13-        run: sed -i "s|\${INSTALL_BCC_TRACING_TOOLS}|true|g" ./ci/test/00_setup_env_native_asan.sh
14-
15       - name: CI script
16         run: ./ci/test_run_all.sh
17

Otherwise, pull requests will unrelatedly and intermittently fail.

maflcko commented at 7:59 am on April 24, 2025: member

Done in 32335 to buy more time to test/fix/workaround this in the meantime.

maflcko added the label Upstream on Apr 24, 2025

hebasto closed this on Apr 24, 2025

hebasto referenced this in commit 458720e5e9 on Apr 24, 2025

ci: failure in interface_usdt_coinselection.py #32322