zfs-2.3.5 patch set

.github/ISSUE_TEMPLATE/feature_request.md

@@ -14,7 +14,7 @@ Please check our issue tracker before opening a new feature request.
 Filling out the following template will help other contributors better understand your proposed feature.
 -->
 
-### Describe the feature would like to see added to OpenZFS
+### Describe the feature you would like to see added to OpenZFS
 
 <!--
 Provide a clear and concise description of the feature.

.github/PULL_REQUEST_TEMPLATE.md

@@ -2,11 +2,6 @@
 
 <!--- Provide a general summary of your changes in the Title above -->
 
-<!---
-Documentation on ZFS Buildbot options can be found at
-https://openzfs.github.io/openzfs-docs/Developer%20Resources/Buildbot%20Options.html
--->
-
 ### Motivation and Context
 <!--- Why is this change required? What problem does it solve? -->
 <!--- If it fixes an open issue, please link to the issue here. -->

.github/workflows/scripts/generate-ci-type.py

@@ -65,7 +65,7 @@ def output_type(type, reason):
 
     # check last (HEAD) commit message
     last_commit_message_raw = subprocess.run([
-        'git', 'show', '-s', '--format=%B', 'HEAD'
+        'git', 'show', '-s', '--format=%B', head
     ], check=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
 
     for line in last_commit_message_raw.stdout.decode().splitlines():

.github/workflows/scripts/qemu-1-setup.sh

@@ -6,6 +6,13 @@
 
 set -eu
 
+# We've been seeing this script take over 15min to run.  This may or
+# may not be normal.  Just to get a little more insight, print out
+# a message to stdout with the top running process, and do this every
+# 30 seconds.  We can delete this watchdog later once we get a better
+# handle on what the timeout value should be.
+(while [ 1 ] ; do sleep 30 && echo "[watchdog: $(ps -eo cmd --sort=-pcpu  | head -n 2 | tail -n 1)}')]"; done) &
+
 # install needed packages
 export DEBIAN_FRONTEND="noninteractive"
 sudo apt-get -y update
@@ -65,3 +72,6 @@ sudo zpool create -f -o ashift=12 zpool $SSD1 $SSD2 -O relatime=off \
 for i in /sys/block/s*/queue/scheduler; do
   echo "none" | sudo tee $i
 done
+
+# Kill off our watchdog
+kill $(jobs -p)

.github/workflows/scripts/qemu-2-start.sh

@@ -121,7 +121,14 @@ case "$OS" in
     KSRC="$FREEBSD_SNAP/../amd64/$FreeBSD/src.txz"
     ;;
   freebsd15-0c)
-    FreeBSD="15.0-PRERELEASE"
+    FreeBSD="15.0-ALPHA4"
+    OSNAME="FreeBSD $FreeBSD"
+    OSv="freebsd14.0"
+    URLxz="$FREEBSD_SNAP/$FreeBSD/amd64/Latest/FreeBSD-$FreeBSD-amd64-BASIC-CI-ufs.raw.xz"
+    KSRC="$FREEBSD_SNAP/../amd64/$FreeBSD/src.txz"
+    ;;
+  freebsd16-0c)
+    FreeBSD="16.0-CURRENT"
     OSNAME="FreeBSD $FreeBSD"
     OSv="freebsd14.0"
     URLxz="$FREEBSD_SNAP/$FreeBSD/amd64/Latest/FreeBSD-$FreeBSD-amd64-BASIC-CI-ufs.raw.xz"
@@ -287,7 +294,7 @@ else
   while pidof /usr/bin/qemu-system-x86_64 >/dev/null; do
     ssh 2>/dev/null root@vm0 "uname -a" && break
   done
-  ssh root@vm0 "pkg install -y bash ca_root_nss git qemu-guest-agent python3 py311-cloud-init"
+  ssh root@vm0 "env IGNORE_OSVERSION=yes pkg install -y bash ca_root_nss git qemu-guest-agent python3 py311-cloud-init"
   ssh root@vm0 "chsh -s $BASH root"
   ssh root@vm0 'sysrc qemu_guest_agent_enable="YES"'
   ssh root@vm0 'sysrc cloudinit_enable="YES"'

.github/workflows/scripts/qemu-3-deps-vm.sh

@@ -20,7 +20,7 @@ function archlinux() {
   sudo pacman -Sy --noconfirm base-devel bc cpio cryptsetup dhclient dkms \
     fakeroot fio gdb inetutils jq less linux linux-headers lsscsi nfs-utils \
     parted pax perf python-packaging python-setuptools qemu-guest-agent ksh \
-    samba sysstat rng-tools rsync wget xxhash
+    samba strace sysstat rng-tools rsync wget xxhash
   echo "##[endgroup]"
 }
 
@@ -43,7 +43,8 @@ function debian() {
     lsscsi nfs-kernel-server pamtester parted python3 python3-all-dev \
     python3-cffi python3-dev python3-distlib python3-packaging libtirpc-dev \
     python3-setuptools python3-sphinx qemu-guest-agent rng-tools rpm2cpio \
-    rsync samba sysstat uuid-dev watchdog wget xfslibs-dev  xxhash zlib1g-dev
+    rsync samba strace sysstat uuid-dev watchdog wget xfslibs-dev xxhash \
+    zlib1g-dev
   echo "##[endgroup]"
 }
 
@@ -87,8 +88,8 @@ function rhel() {
     libuuid-devel lsscsi mdadm nfs-utils openssl-devel pam-devel pamtester \
     parted perf python3 python3-cffi python3-devel python3-packaging \
     kernel-devel python3-setuptools qemu-guest-agent rng-tools rpcgen \
-    rpm-build rsync samba sysstat systemd watchdog wget xfsprogs-devel xxhash \
-    zlib-devel
+    rpm-build rsync samba strace sysstat systemd watchdog wget xfsprogs-devel \
+    xxhash zlib-devel
   echo "##[endgroup]"
 }
 
@@ -104,7 +105,7 @@ function install_fedora_experimental_kernel {
   our_version="$1"
   sudo dnf -y copr enable @kernel-vanilla/stable
   sudo dnf -y copr enable @kernel-vanilla/mainline
-  all="$(sudo dnf list --showduplicates kernel-*)"
+  all="$(sudo dnf list --showduplicates kernel-* python3-perf* perf* bpftool*)"
   echo "Available versions:"
   echo "$all"
 

.github/workflows/scripts/qemu-5-setup.sh

@@ -108,19 +108,30 @@ echo '*/5 * * * *  /root/cronjob.sh' > crontab.txt
 sudo crontab crontab.txt
 rm crontab.txt
 
-# check if the machines are okay
-echo "Waiting for vm's to come up...  (${VMs}x CPU=$CPU RAM=$RAM)"
-for ((i=1; i<=VMs; i++)); do
-  .github/workflows/scripts/qemu-wait-for-vm.sh vm$i
-done
-echo "All $VMs VMs are up now."
-
 # Save the VM's serial output (ttyS0) to /var/tmp/console.txt
 # - ttyS0 on the VM corresponds to a local /dev/pty/N entry
 # - use 'virsh ttyconsole' to lookup the /dev/pty/N entry
 for ((i=1; i<=VMs; i++)); do
   mkdir -p $RESPATH/vm$i
   read "pty" <<< $(sudo virsh ttyconsole vm$i)
+
+  # Create the file so we can tail it, even if there's no output.
+  touch $RESPATH/vm$i/console.txt
+
   sudo nohup bash -c "cat $pty > $RESPATH/vm$i/console.txt" &
+
+  # Write all VM boot lines to the console to aid in debugging failed boots.
+  # The boot lines from all the VMs will be munged together, so prepend each
+  # line with the vm hostname (like 'vm1:').
+  (while IFS=$'\n' read -r line; do echo "vm$i: $line" ; done < <(sudo tail -f $RESPATH/vm$i/console.txt)) &
+
 done
 echo "Console logging for ${VMs}x $OS started."
+
+
+# check if the machines are okay
+echo "Waiting for vm's to come up...  (${VMs}x CPU=$CPU RAM=$RAM)"
+for ((i=1; i<=VMs; i++)); do
+  .github/workflows/scripts/qemu-wait-for-vm.sh vm$i
+done
+echo "All $VMs VMs are up now."

.github/workflows/scripts/qemu-6-tests.sh

@@ -111,7 +111,7 @@ fi
 sudo dmesg -c > dmesg-prerun.txt
 mount > mount.txt
 df -h > df-prerun.txt
-$TDIR/zfs-tests.sh -vK -s 3GB -T $TAGS
+$TDIR/zfs-tests.sh -vKO -s 3GB -T $TAGS
 RV=$?
 df -h > df-postrun.txt
 echo $RV > tests-exitcode.txt

.github/workflows/zfs-qemu.yml

@@ -44,7 +44,7 @@ jobs:
             os_selection="$FULL_OS"
           fi
 
-          if [ ${{ github.event.inputs.fedora_kernel_ver }} != "" ] ; then
+          if ${{ github.event.inputs.fedora_kernel_ver != '' }}; then
               # They specified a custom kernel version for Fedora.  Use only
               # Fedora runners.
               os_json=$(echo ${os_selection} | jq -c '[.[] | select(startswith("fedora"))]')
@@ -53,9 +53,8 @@ jobs:
               os_json=$(echo ${os_selection} | jq -c)
           fi
 
-          echo $os_json
-          echo "os=$os_json" >> $GITHUB_OUTPUT
-          echo "ci_type=$ci_type" >> $GITHUB_OUTPUT
+          echo "os=$os_json" | tee -a $GITHUB_OUTPUT
+          echo "ci_type=$ci_type" | tee -a $GITHUB_OUTPUT
 
   qemu-vm:
     name: qemu-x86
@@ -69,7 +68,7 @@ jobs:
         # FreeBSD variants of 2025-06:
         # FreeBSD Release: freebsd13-5r, freebsd14-2r, freebsd14-3r
         # FreeBSD Stable:  freebsd13-5s, freebsd14-3s
-        # FreeBSD Current: freebsd15-0c
+        # FreeBSD Current: freebsd15-0c, freebsd16-0c
         os: ${{ fromJson(needs.test-config.outputs.test_os) }}
     runs-on: ubuntu-24.04
     steps:
@@ -78,8 +77,12 @@ jobs:
         ref: ${{ github.event.pull_request.head.sha }}
 
     - name: Setup QEMU
-      timeout-minutes: 10
-      run: .github/workflows/scripts/qemu-1-setup.sh
+      timeout-minutes: 20
+      run: |
+        # Add a timestamp to each line to debug timeouts
+        while IFS=$'\n' read -r line; do
+            echo "$(date +'%H:%M:%S') $line"
+        done < <(.github/workflows/scripts/qemu-1-setup.sh)
 
     - name: Start build machine
       timeout-minutes: 10

META

@@ -1,10 +1,10 @@
 Meta:          1
 Name:          zfs
 Branch:        1.0
-Version:       2.3.4
+Version:       2.3.5
 Release:       1
 Release-Tags:  relext
 License:       CDDL
 Author:        OpenZFS
-Linux-Maximum: 6.16
+Linux-Maximum: 6.17
 Linux-Minimum: 4.18

cmd/zdb/zdb.c

@@ -381,7 +381,7 @@ verify_livelist_allocs(metaslab_verify_t *mv, uint64_t txg,
 	sublivelist_verify_block_t svb = {{{0}}};
 	DVA_SET_VDEV(&svb.svb_dva, mv->mv_vdid);
 	DVA_SET_OFFSET(&svb.svb_dva, offset);
-	DVA_SET_ASIZE(&svb.svb_dva, size);
+	DVA_SET_ASIZE(&svb.svb_dva, 0);
 	zfs_btree_index_t where;
 	uint64_t end_offset = offset + size;
 

cmd/zfs/zfs_main.c

@@ -930,19 +930,15 @@ zfs_do_clone(int argc, char **argv)
 }
 
 /*
- * Return a default volblocksize for the pool which always uses more than
- * half of the data sectors.  This primarily applies to dRAID which always
- * writes full stripe widths.
+ * Calculate the minimum allocation size based on the top-level vdevs.
  */
 static uint64_t
-default_volblocksize(zpool_handle_t *zhp, nvlist_t *props)
+calculate_volblocksize(nvlist_t *config)
 {
-	uint64_t volblocksize, asize = SPA_MINBLOCKSIZE;
+	uint64_t asize = SPA_MINBLOCKSIZE;
 	nvlist_t *tree, **vdevs;
 	uint_t nvdevs;
 
-	nvlist_t *config = zpool_get_config(zhp, NULL);
-
 	if (nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, &tree) != 0 ||
 	    nvlist_lookup_nvlist_array(tree, ZPOOL_CONFIG_CHILDREN,
 	    &vdevs, &nvdevs) != 0) {
@@ -973,6 +969,24 @@ default_volblocksize(zpool_handle_t *zhp, nvlist_t *props)
 		}
 	}
 
+	return (asize);
+}
+
+/*
+ * Return a default volblocksize for the pool which always uses more than
+ * half of the data sectors.  This primarily applies to dRAID which always
+ * writes full stripe widths.
+ */
+static uint64_t
+default_volblocksize(zpool_handle_t *zhp, nvlist_t *props)
+{
+	uint64_t volblocksize, asize = SPA_MINBLOCKSIZE;
+
+	nvlist_t *config = zpool_get_config(zhp, NULL);
+
+	if (nvlist_lookup_uint64(config, ZPOOL_CONFIG_MAX_ALLOC, &asize) != 0)
+		asize = calculate_volblocksize(config);
+
 	/*
 	 * Calculate the target volblocksize such that more than half
 	 * of the asize is used. The following table is for 4k sectors.

cmd/zpool/zpool_vdev.c

@@ -609,23 +609,29 @@ get_replication(nvlist_t *nvroot, boolean_t fatal)
 				verify(nvlist_lookup_string(cnv,
 				    ZPOOL_CONFIG_PATH, &path) == 0);
 
+				/*
+				 * Skip active spares they should never cause
+				 * the pool to be evaluated as inconsistent.
+				 */
+				if (is_spare(NULL, path))
+					continue;
+
 				/*
 				 * If we have a raidz/mirror that combines disks
-				 * with files, report it as an error.
+				 * with files, only report it as an error when
+				 * fatal is set to ensure all the replication
+				 * checks aren't skipped in check_replication().
 				 */
-				if (!dontreport && type != NULL &&
+				if (fatal && !dontreport && type != NULL &&
 				    strcmp(type, childtype) != 0) {
 					if (ret != NULL)
 						free(ret);
 					ret = NULL;
-					if (fatal)
-						vdev_error(gettext(
-						    "mismatched replication "
-						    "level: %s contains both "
-						    "files and devices\n"),
-						    rep.zprl_type);
-					else
-						return (NULL);
+					vdev_error(gettext(
+					    "mismatched replication "
+					    "level: %s contains both "
+					    "files and devices\n"),
+					    rep.zprl_type);
 					dontreport = B_TRUE;
 				}
 

config/kernel-dentry-operations.m4

@@ -24,6 +24,9 @@ dnl #
 dnl # 2.6.38 API change
 dnl # Added d_set_d_op() helper function.
 dnl #
+dnl # 6.17 API change
+dnl # d_set_d_op() removed. No direct replacement.
+dnl #
 AC_DEFUN([ZFS_AC_KERNEL_SRC_D_SET_D_OP], [
 	ZFS_LINUX_TEST_SRC([d_set_d_op], [
 		#include <linux/dcache.h>
@@ -34,11 +37,12 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_D_SET_D_OP], [
 
 AC_DEFUN([ZFS_AC_KERNEL_D_SET_D_OP], [
 	AC_MSG_CHECKING([whether d_set_d_op() is available])
-	ZFS_LINUX_TEST_RESULT_SYMBOL([d_set_d_op],
-	    [d_set_d_op], [fs/dcache.c], [
+	ZFS_LINUX_TEST_RESULT([d_set_d_op], [
 		AC_MSG_RESULT(yes)
+		AC_DEFINE(HAVE_D_SET_D_OP, 1,
+		    [Define if d_set_d_op() is available])
 	], [
-		ZFS_LINUX_TEST_ERROR([d_set_d_op])
+		AC_MSG_RESULT(no)
 	])
 ])
 

contrib/initramfs/scripts/zfs

@@ -979,7 +979,8 @@ mountroot()
 
 	touch /run/zfs_unlock_complete
 	if [ -e /run/zfs_unlock_complete_notify ]; then
-		read -r < /run/zfs_unlock_complete_notify
+		# shellcheck disable=SC2034
+		read -r zfs_unlock_complete_notify < /run/zfs_unlock_complete_notify
 	fi
 
 	# ------------

contrib/intel_qat/readme.md

@@ -8,7 +8,7 @@ This contrib contains community compatibility patches to get Intel QAT working o
 These patches are based on the following Intel QAT version:
 [1.7.l.4.10.0-00014](https://01.org/sites/default/files/downloads/qat1.7.l.4.10.0-00014.tar.gz)
 
-When using QAT with above kernels versions, the following patches needs to be applied using:
+When using QAT with the above kernel versions, the following patches need to be applied using:
 patch -p1 < _$PATCH_
 _Where $PATCH refers to the path of the patch in question_
 

contrib/pyzfs/libzfs_core/test/test_libzfs_core.py

@@ -4223,7 +4223,7 @@ def reset(self):
             self.getRoot().reset()
             return
 
-        # On the Buildbot builders this may fail with "pool is busy"
+        # On the CI builders this may fail with "pool is busy"
         # Retry 5 times before raising an error
         retry = 0
         while True:

etc/init.d/README.md

@@ -1,5 +1,5 @@
 DESCRIPTION
-  These script were written with the primary intention of being portable and
+  These scripts were written with the primary intention of being portable and
   usable on as many systems as possible.
 
   This is, in practice, usually not possible. But the intention is there.