Patch fakeroot in buildroot to avoid build hanging #22432

afbjorklund · 2026-01-10T17:25:26Z

When running in docker, there could be a large range of file descriptors causing "faked" to hang at the end...

Apply the same workaround for this as was done in apt, as posted on the mailing list. Debian Bug no: 920913

EDIT: Seems like this issue was reported to buildroot too:

https://gitlab.com/buildroot.org/buildroot/-/issues/115

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=920913;msg=25

I rebased it to fakeroot 1.36, and minimized the patch by ignoring whitespace.

Fixes the issue when you might have increased the limits, on a real Linux host.

The official "fix" in 1.35 only adds close_range, which still takes too long time...

When running in docker, there could be a large range of file descriptors causing "faked" to hang at the end. Apply the same workaround for this as was done in apt, as posted on the mailing list. Debian Bug no: 920913

k8s-ci-robot · 2026-01-10T17:25:33Z

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: afbjorklund
Once this PR has been reviewed and has the lgtm label, please assign comradeprogrammer for approval. For more information see the Code Review Process.

The full list of commands accepted by this bot can be found here.

Details

Needs approval from an approver in each of these files:

OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

medyagh · 2026-01-11T06:47:25Z

/ok-to-build-iso

Copilot

Pull request overview

This PR adds a patch for fakeroot 1.36 to optimize file descriptor closing in containers where ulimit may be very high. The patch implements an optimization that iterates through /proc/self/fd to close only open file descriptors, instead of blindly iterating through all possible file descriptor numbers up to the ulimit.

Changes:

Added a new patch file for fakeroot 1.36 that implements fd-closing optimization based on the approach used in APT
The patch adds logic to read actual open file descriptors from /proc/self/fd before falling back to the original approach

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

deploy/iso/minikube-iso/patches/fakeroot/1.36/optimize-fd-closing.patch

Copilot · 2026-01-11T06:49:45Z

deploy/iso/minikube-iso/patches/fakeroot/1.36/optimize-fd-closing.patch

+      if (dir != NULL) {
+        struct dirent *ent;
+        int dfd = dirfd(dir);
+        while (ent = readdir(dir)) {


The while loop assignment lacks parentheses around the assignment expression. The C compiler will issue a warning for this pattern. The condition should be while ((ent = readdir(dir))) instead of while (ent = readdir(dir)) to make the assignment-as-condition pattern explicit and suppress compiler warnings.

Suggested change

+ while (ent = readdir(dir)) {

+ while ((ent = readdir(dir)) != NULL) {

Copilot · 2026-01-11T06:49:45Z

deploy/iso/minikube-iso/patches/fakeroot/1.36/optimize-fd-closing.patch

+          fl = atoi(ent->d_name);
+          if (fl >= 0 && fl != dfd)


The use of atoi on directory entry names is problematic. When readdir encounters "." or ".." entries, atoi will return 0, which would incorrectly attempt to close stdin (fd 0). Additionally, fl >= 0 allows closing fd 0. The condition should check fl > 0 to avoid closing stdin, and should verify the entry name consists only of digits before calling atoi to avoid issues with non-numeric directory entries.

Suggested change

+ fl = atoi(ent->d_name);

+ if (fl >= 0 && fl != dfd)

+ const char *name = ent->d_name;

+ int is_numeric = 1;

+ const char *p;

+

+ for (p = name; *p != '\0'; ++p) {

+ if (!isdigit((unsigned char)*p)) {

+ is_numeric = 0;

+ break;

+ }

+ }

+ if (!is_numeric)

+ continue;

+

+ fl = atoi(name);

+ if (fl > 0 && fl != dfd)

Copilot · 2026-01-11T06:49:46Z

deploy/iso/minikube-iso/patches/fakeroot/1.36/optimize-fd-closing.patch

+          if (fl >= 0 && fl != dfd)
+#ifdef FAKEROOT_FAKENET
+            if (fl != sd)
+#endif /* FAKEROOT_FAKENET */
+              close(fl);


The control flow structure with nested preprocessor conditionals creates ambiguous logic. When FAKEROOT_FAKENET is defined, there are two consecutive if statements without proper bracing, making it unclear which conditions must be met for the close call to execute. This should use proper logical operators (&&) or explicit bracing to clarify that both conditions (fl != dfd AND fl != sd) must be true before closing the file descriptor.

Suggested change

+ if (fl >= 0 && fl != dfd)

+#ifdef FAKEROOT_FAKENET

+ if (fl != sd)

+#endif /* FAKEROOT_FAKENET */

+ close(fl);

+ if (fl >= 0 && fl != dfd

+#ifdef FAKEROOT_FAKENET

+ && fl != sd

+#endif /* FAKEROOT_FAKENET */

+ )

+ close(fl);

afbjorklund · 2026-01-11T09:08:08Z

This 2021 patch is not yet accepted upstream, when it is it might be more worthwhile to spend time on "fixing" it there.

https://salsa.debian.org/clint/fakeroot (and https://bugs.debian.org/cgi-bin/pkgreport.cgi?src=fakeroot)

minikube-bot · 2026-01-11T09:33:21Z

Hi @afbjorklund, building a new ISO failed for Commit b2c401f
See the logs at:
https://storage.cloud.google.com/minikube-builds/logs/22432/b2c401f/iso_build.txt

afbjorklund · 2026-01-11T10:59:21Z

Fakeroot package built OK, then Jenkins ran out of disk or something:

/home/jenkins/workspace/iso-pr-build/out/buildroot/output-x86_64/host/bin/tar: /home/jenkins/workspace/iso-pr-build/out/buildroot/output-x86_64/images/rootfs.tar: Wrote only 6144 of 10240 bytes

medyagh · 2026-01-17T07:22:56Z

/ok-to-build-iso

medyagh · 2026-01-17T07:23:20Z

@afbjorklund do you mind checking the copilot's comments to see if they are relevant ?

afbjorklund · 2026-01-17T08:33:36Z

I did (#22432 (comment)), but I'm not sure I want to go about "fixing" whitespace in every patch

minikube-bot · 2026-01-17T10:09:44Z

Hi @afbjorklund, we have updated your PR with the reference to newly built ISO. Pull the changes locally if you want to test with them or update your PR further.

k8s-ci-robot · 2026-01-17T10:09:52Z

PR needs rebase.

Details

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

k8s-ci-robot · 2026-01-17T10:13:35Z

@afbjorklund: The following tests failed, say /retest to rerun all failed tests or /retest-required to rerun all mandatory failed tests:

Test name	Commit	Details	Required	Rerun command
integration-docker-containerd-linux-x86	`a4c91ad`	link	true	`/test integration-docker-containerd-linux-x86`
integration-kvm-crio-linux-x86	`a4c91ad`	link	false	`/test integration-kvm-crio-linux-x86`
integration-kvm-containerd-linux-x86	`a4c91ad`	link	true	`/test integration-kvm-containerd-linux-x86`
pull-minikube-build	`a4c91ad`	link	true	`/test pull-minikube-build`
integration-docker-docker-linux-x86	`a4c91ad`	link	true	`/test integration-docker-docker-linux-x86`
integration-kvm-docker-linux-x86	`a4c91ad`	link	true	`/test integration-kvm-docker-linux-x86`
integration-docker-crio-linux-x86	`a4c91ad`	link	false	`/test integration-docker-crio-linux-x86`
integration-none-containerd-linux-x86	`a4c91ad`	link	false	`/test integration-none-containerd-linux-x86`
integration-none-docker-linux-x86	`a4c91ad`	link	true	`/test integration-none-docker-linux-x86`
integration-docker-containerd-linux-arm	`a4c91ad`	link	false	`/test integration-docker-containerd-linux-arm`
integration-vfkit-docker-macos-arm	`a4c91ad`	link	false	`/test integration-vfkit-docker-macos-arm`
integration-docker-docker-linux-arm	`a4c91ad`	link	true	`/test integration-docker-docker-linux-arm`

Full PR test history. Your PR dashboard. Please help us cut down on flakes by linking to an open issue when you hit one in your PR.

Details

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. I understand the commands that are listed here.

Patch fakeroot in buildroot to avoid build hanging

b2c401f

When running in docker, there could be a large range of file descriptors causing "faked" to hang at the end. Apply the same workaround for this as was done in apt, as posted on the mailing list. Debian Bug no: 920913

k8s-ci-robot requested a review from ComradeProgrammer January 10, 2026 17:25

k8s-ci-robot requested a review from nirs January 10, 2026 17:25

k8s-ci-robot added cncf-cla: yes Indicates the PR's author has signed the CNCF CLA. size/M Denotes a PR that changes 30-99 lines, ignoring generated files. labels Jan 10, 2026

afbjorklund requested a review from medyagh January 10, 2026 17:25

medyagh requested a review from Copilot January 11, 2026 06:47

Copilot started reviewing on behalf of medyagh January 11, 2026 06:47 View session

Copilot AI reviewed Jan 11, 2026

View reviewed changes

afbjorklund mentioned this pull request Jan 13, 2026

iso: Add linux tarball hash for custom linux version #22436

Merged

Updating ISO to v1.37.0-1768634666-22432

a4c91ad

k8s-ci-robot added the needs-rebase Indicates a PR cannot be merged because it has merge conflicts with HEAD. label Jan 17, 2026

	+ while (ent = readdir(dir)) {
	+ while ((ent = readdir(dir)) != NULL) {

-+          fl = atoi(ent->d_name);
-+          if (fl >= 0 && fl != dfd)
++          const char *name = ent->d_name;
++          int is_numeric = 1;
++          const char *p;
++
++          for (p = name; *p != '\0'; ++p) {
++            if (!isdigit((unsigned char)*p)) {
++              is_numeric = 0;
++              break;
++            }
++          }
++          if (!is_numeric)
++            continue;
++
++          fl = atoi(name);
++          if (fl > 0 && fl != dfd)

Patch fakeroot in buildroot to avoid build hanging #22432

Are you sure you want to change the base?

Patch fakeroot in buildroot to avoid build hanging #22432

Conversation

afbjorklund commented Jan 10, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

k8s-ci-robot commented Jan 10, 2026

Uh oh!

medyagh commented Jan 11, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Copilot AI Jan 11, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 11, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 11, 2026

Choose a reason for hiding this comment

Uh oh!

afbjorklund commented Jan 11, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

minikube-bot commented Jan 11, 2026

Uh oh!

afbjorklund commented Jan 11, 2026

Uh oh!

medyagh commented Jan 17, 2026

Uh oh!

medyagh commented Jan 17, 2026

Uh oh!

afbjorklund commented Jan 17, 2026

Uh oh!

minikube-bot commented Jan 17, 2026

Uh oh!

k8s-ci-robot commented Jan 17, 2026

Uh oh!

k8s-ci-robot commented Jan 17, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

afbjorklund commented Jan 10, 2026 •

edited

Loading

afbjorklund commented Jan 11, 2026 •

edited

Loading