Hi John,
On Tue, Nov 25, 2025 at 03:30:43PM +0100, John Paul Adrian Glaubitz wrote:
Hi Bo,
On Tue, 2025-11-25 at 22:02 +0800, Bo YU wrote:
The package has one FTBFS issue on riscv64 like:
(...)
See the full buildd log:
https://buildd.debian.org/status/fetch.php?pkg=fish&arch=riscv64&ver=4.2.1-2&stamp=1764008323&raw=0
I admit the timeout I increased were too excessive for some cases. Even so,
the build with the patch is not 100% successful. But I have confidence
to get it built on buildd. Please let me know any issue.
I think your patch name and changelog entry is misleading as your change
affects all architectures.
You are right. Here I did not filter out riscv64 here.
I would rather phrase it "Increase timeouts to fix FTBFS on riscv64".
Okay, thanks for the improvement.
And while you're at it, it would be nice if you could increase the time for
tests/pexpects/fkr.py
as well as it times out on sparc64 [1].
Done.
Becase I do not test it on sparc64 but I assume it also has different random timeout issues as riscv64
within the same test unit, maybe I overdo it.
Thanks,
Adrian
--
Regards,
--
Bo YU
diff -Nru fish-4.2.1/debian/changelog fish-4.2.1/debian/changelog
--- fish-4.2.1/debian/changelog 2025-11-22 00:40:34.0 +0800
+++ fish-4.2.1/debian/changelog 2025-11-25 17:56:47.0 +0800
@@ -1,3 +1,10 @@
+fish (4.2.1-2.1) UNRELEASED; urgency=medium
+
+ * Non-maintainer upload.
+ * Increase timeouts to fix FTBFS on riscv64 and sparc64. (Closes: #1121377)
+
+ -- Bo YU Tue, 25 Nov 2025 17:56:47 +0800
+
fish (4.2.1-2) unstable; urgency=medium
* Reinstate fix-libc-timespec.patch for armhf
diff -Nru fish-4.2.1/debian/patches/increase-timeout-on-slow-architecture.patch
fish-4.2.1/debian/patches/increase-timeout-on-slow-architecture.patch
--- fish-4.2.1/debian/patches/increase-timeout-on-slow-architecture.patch
1970-01-01 08:00:00.0 +0800
+++ fish-4.2.1/debian/patches/increase-timeout-on-slow-architecture.patch
2025-11-25 17:56:47.0 +0800
@@ -0,0 +1,116 @@
+Description: Increase timeouts to fix FTBFS on riscv64 and sparc64
+Author: Bo YU
+Bug: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1121377#10
+Last-Update: 2025-11-25
+---
+This patch header follows DEP-3: http://dep.debian.net/deps/dep3/
+--- a/tests/pexpects/complete.py
b/tests/pexpects/complete.py
+@@ -18,16 +18,16 @@
+ sendline("complete -x -c monster -a energy=")
+ expect_prompt()
+ send("monster t")
+-sleep(0.1)
++sleep(10.1)
+ send("\t")
+-sleep(0.1)
++sleep(10.1)
+ send("!")
+ expect_str("monster truck !") # space
+ send("\b" * 64)
+ send("monster e")
+-sleep(0.1)
++sleep(10.1)
+ send("\t")
+-sleep(0.1)
++sleep(10.1)
+ send("!")
+ expect_str("monster energy=!") # no space
+ send("\b" * 64)
+--- a/tests/pexpects/job_summary.py
b/tests/pexpects/job_summary.py
+@@ -30,7 +30,7 @@
+ # fish_job_summary is called when background job ends.
+ sendline("sleep 0.5 &")
+ expect_prompt()
+-expect_re("[0-9]+:0:sleep 0.5 &:ENDED", timeout=20)
++expect_re("[0-9]+:0:sleep 0.5 &:ENDED", timeout=40)
+ sendline("")
+ expect_prompt()
+
+@@ -48,7 +48,7 @@
+ m = expect_re("\\d+\r\n")
+ expect_prompt()
+ os.kill(int(m.group()), signal.SIGTERM)
+-expect_re("[0-9]+:0:sleep 20 &:SIGTERM:Polite quit request", timeout=20)
++expect_re("[0-9]+:0:sleep 20 &:SIGTERM:Polite quit request", timeout=40)
+ sendline("")
+ expect_prompt()
+
+--- a/tests/checks/tmux-job.fish
b/tests/checks/tmux-job.fish
+@@ -8,9 +8,9 @@
+ sleep 0.1
+ isolated-tmux send-keys \
+ "echo hello"
+-sleep 0.6
++sleep 5.6
+ isolated-tmux send-keys Space world
+-sleep 0.1
++sleep 5.1
+ isolated-tmux capture-pane -p
+ # CHECK: prompt 0> sleep 0.5 &
+ # CHECK: prompt 0> echo hello
+--- a/tests/checks/tmux-prompt.fish
b/tests/checks/tmux-prompt.fish
+@@ -51,6 +51,7 @@
+ end
+ ' C-l 'echo hello'
+ tmux-sleep
++tmux-sleep
+ isolated-tmux capture-pane -p
+ # CHECK:
+ # CHECK: echo hello
+@@ -64,6 +65,8 @@
+ end
+ ' Enter
+ tmux-sleep
++tmux-sleep
++tmux-sleep
+ isolated-tmux capture-pane -p -S -11
+ # CHECK: 1
+ # CHECK: 2
+--- a/tests/pexpects/fkr.py
b/tests/pexpects/fkr.py
+@@ -35,7 +35,7 @@
+ expect_str("bind ctrl-g 'do something'\r\n")
+
+ # Is a non-ASCII UTF-8 sequence prefaced by an escape char handled correctly?
+-sleep(0.020)
++sleep(1.020)
+ send("\x1b")
+ expect_str("# decoded from: \\e\r\n")
+ expect_str("bind escape 'do something'\r\n")
+@@ -46,7 +46,7 @@
+ expect_str("bind ሴ 'do something'\r\n")
+
+ # Is a NULL char echoed correctly?
+-sleep(0.020)
++sleep(1.020)
+ send("\x00")
+ expect_str("bind ctrl-space 'do something'\r\n")
+
+@@ -59,11 +59,11 @@
+
+ # Does it keep running if handed control sequences in the wrong order?
+ send("\x03")
+-sleep(0.010)
++sleep(1.010)
+ send("\x04")
+
+ # Now send a second ctrl-d. Does that terminate the process like it should?
+-sleep(0.050)
++sleep(1.050)
+ send("\x04")
