[PATCH 00/10] More test fixes/improvements

[PATCH 00/10] More test fixes/improvements

[Glenn Washburn]

Here is another patch series of changes needed to get the make check tests
passing on fairly recent Ubuntu/Debian systems, though most of the fixes
are for issues on fairly old systems as well (eg. reiser fixes). The FAT
volume label change is only needed on fairly recent systems because up until
very recently a fairly old version of dosfstools has been used which didn't
do invalid character checking.

Glenn

Glenn Washburn (10):
  tests: Rename variable filtime -> filetime as its meant to be
  tests: mkreiserfs only supports 4096 block size
  tests: Disable reiserfs tests for old format because newer kernels do
    not support them
  tests: mkfs.btrfs now supports only 4k sector sizes and above
  tests: Only test minix3 volumes of 1k block size
  tests: Change FAT volume label to be with in the valid character range
  tests: Skip hfs test only when mac_roman module is not loaded and not
    loadable
  grub-shell: Boot PowerPC using PMU instead of CUDA for power
    management
  tests: Output list of devices when partmap fails
  tests: Do not delete filesystem images on error

 tests/hfs_test.in            |  2 +-
 tests/minixfs_test.in        |  5 -----
 tests/partmap_test.in        |  2 +-
 tests/reiserfs_test.in       |  4 +++-
 tests/util/grub-fs-tester.in | 28 ++++++++++++++--------------
 tests/util/grub-shell.in     |  1 +
 6 files changed, 20 insertions(+), 22 deletions(-)

-- 
2.32.0

[PATCH 01/10] tests: Rename variable filtime -> filetime as its meant to be

[Glenn Washburn]

Signed-off-by: Glenn Washburn <development@efficientek.com>
---
 tests/util/grub-fs-tester.in | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/tests/util/grub-fs-tester.in b/tests/util/grub-fs-tester.in
index 7c22cf882..9f6f26f26 100644
--- a/tests/util/grub-fs-tester.in
+++ b/tests/util/grub-fs-tester.in
@@ -1170,8 +1170,8 @@ for LOGSECSIZE in $(range "$MINLOGSECSIZE" "$MAXLOGSECSIZE" 1); do
 	    fi
 
 	    if [ x$NOFILETIME != xy ]; then
-		filtime=$(TZ=UTC ls --time-style=+%Y%m%d%H%M%S -l -d "$MNTPOINTRO/$OSDIR/$BASEFILE"|awk '{print $6; }')
-		if echo "$LSROUT" | grep -F "$filtime $BASEFILE" > /dev/null; then
+		filetime=$(TZ=UTC ls --time-style=+%Y%m%d%H%M%S -l -d "$MNTPOINTRO/$OSDIR/$BASEFILE"|awk '{print $6; }')
+		if echo "$LSROUT" | grep -F "$filetime $BASEFILE" > /dev/null; then
 		    :
 		else
 		    echo TIME FAIL
@@ -1180,8 +1180,8 @@ for LOGSECSIZE in $(range "$MINLOGSECSIZE" "$MAXLOGSECSIZE" 1); do
 		    exit 1
 		fi
 
-		filtime=$(TZ=UTC ls --time-style=+%Y%m%d%H%M%S -l -d "$MNTPOINTRO/$OSDIR/$LONGNAME"|awk '{print $6; }')
-		if echo "$LSROUT" | grep -F "$filtime $LONGNAME" > /dev/null; then
+		filetime=$(TZ=UTC ls --time-style=+%Y%m%d%H%M%S -l -d "$MNTPOINTRO/$OSDIR/$LONGNAME"|awk '{print $6; }')
+		if echo "$LSROUT" | grep -F "$filetime $LONGNAME" > /dev/null; then
 		    :
 		else
 		    echo LONG TIME FAIL
-- 
2.32.0

[PATCH 02/10] tests: mkreiserfs only supports 4096 block size

[Glenn Washburn]

Signed-off-by: Glenn Washburn <development@efficientek.com>
---
 tests/util/grub-fs-tester.in | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/tests/util/grub-fs-tester.in b/tests/util/grub-fs-tester.in
index 9f6f26f26..ab58a96c8 100644
--- a/tests/util/grub-fs-tester.in
+++ b/tests/util/grub-fs-tester.in
@@ -126,8 +126,10 @@ for LOGSECSIZE in $(range "$MINLOGSECSIZE" "$MAXLOGSECSIZE" 1); do
 	    MAXBLKSIZE=4096
 	    ;;
 	xreiserfs*)
-	    MINBLKSIZE=512
-		# OS LIMITATION: 8192 isn't supported.
+	    # OS LIMITATION: mkreiserfs says block sizes smaller than 4k are
+	    # not supported and bails. Its been this way since at least 2014.
+	    MINBLKSIZE=4096
+	    # OS LIMITATION: 8192 isn't supported.
 	    MAXBLKSIZE=4096
 	    ;;
 	x"mdraid"*)
-- 
2.32.0

[PATCH 03/10] tests: Disable reiserfs tests for old format because newer kernels do not support them

[Glenn Washburn]

Signed-off-by: Glenn Washburn <development@efficientek.com>
---
 tests/reiserfs_test.in | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/tests/reiserfs_test.in b/tests/reiserfs_test.in
index b5fed7635..ff703b203 100644
--- a/tests/reiserfs_test.in
+++ b/tests/reiserfs_test.in
@@ -16,5 +16,7 @@ if ! which mkfs.reiserfs >/dev/null 2>&1; then
 fi
 
 "@builddir@/grub-fs-tester" reiserfs
-"@builddir@/grub-fs-tester" reiserfs_old
+
+# Kernels since at least 4.15 can not mount reiser filesystems of the old format
+#"@builddir@/grub-fs-tester" reiserfs_old
 
-- 
2.32.0

[PATCH 04/10] tests: mkfs.btrfs now supports only 4k sector sizes and above

[Glenn Washburn]

Signed-off-by: Glenn Washburn <development@efficientek.com>
---
 tests/util/grub-fs-tester.in | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/util/grub-fs-tester.in b/tests/util/grub-fs-tester.in
index ab58a96c8..f4d9f161b 100644
--- a/tests/util/grub-fs-tester.in
+++ b/tests/util/grub-fs-tester.in
@@ -59,7 +59,7 @@ case x"$fs" in
 	MINLOGSECSIZE=8
 	MAXLOGSECSIZE=12;;
     xbtrfs*)
-	MINLOGSECSIZE=8
+	MINLOGSECSIZE=12
 	    #  OS LIMITATION: It could go up to 32768 but Linux rejects sector sizes > 4096
 	MAXLOGSECSIZE=12;;
     xxfs)
-- 
2.32.0

[PATCH 05/10] tests: Only test minix3 volumes of 1k block size

[Glenn Washburn]

Apparently there used to be a -B option for mkfs.minix to create a volume
with a specified block size. This version is hard to come by and does not
appear to be available in debian distributions. So remove support for
testing a variety of blocks sizes for minix3. This allows the minix tests
to run because they were being skipped due to not finding a mkfs.minix with
the -B option.

Signed-off-by: Glenn Washburn <development@efficientek.com>
---
 tests/minixfs_test.in        | 5 -----
 tests/util/grub-fs-tester.in | 4 ++--
 2 files changed, 2 insertions(+), 7 deletions(-)

diff --git a/tests/minixfs_test.in b/tests/minixfs_test.in
index 3b16a4de0..437d92df6 100644
--- a/tests/minixfs_test.in
+++ b/tests/minixfs_test.in
@@ -25,11 +25,6 @@ if ! mkfs.minix -h | grep -- -3 > /dev/null; then
    exit 77
 fi
 
-if ! mkfs.minix -h | grep -- -B > /dev/null; then
-    echo "mkfs.minix doesn't support variable block size; cannot test minix*fs."
-   exit 77
-fi
-
 "@builddir@/grub-fs-tester" minix
 "@builddir@/grub-fs-tester" minix2
 "@builddir@/grub-fs-tester" minix3
diff --git a/tests/util/grub-fs-tester.in b/tests/util/grub-fs-tester.in
index f4d9f161b..96b204590 100644
--- a/tests/util/grub-fs-tester.in
+++ b/tests/util/grub-fs-tester.in
@@ -201,7 +201,7 @@ for LOGSECSIZE in $(range "$MINLOGSECSIZE" "$MAXLOGSECSIZE" 1); do
 		# OS LIMITATION: Linux rejects non-power-of-two blocks.
 		# OS LIMITATION: Linux rejects > 4096.
 	    MINBLKSIZE=1024
-	    MAXBLKSIZE=4096;;
+	    MAXBLKSIZE=1024;;
     esac
     if test "$BLKSTEP" -eq 0; then
 	blksizes="$(powrange "$MINBLKSIZE" "$MAXBLKSIZE")"
@@ -697,7 +697,7 @@ for LOGSECSIZE in $(range "$MINLOGSECSIZE" "$MAXLOGSECSIZE" 1); do
 		    "mkfs.minix" -v "${MOUNTDEVICE}"
 		    MOUNTFS="minix";;
 		x"minix3")
-		    "mkfs.minix" -B $BLKSIZE -3 "${MOUNTDEVICE}"
+		    "mkfs.minix" -3 "${MOUNTDEVICE}"
 		    MOUNTFS="minix";;
 		x"ntfs"*)
 		    "mkfs.ntfs" -s "$SECSIZE" -c "$BLKSIZE" -L "$FSLABEL" -Q -q "${MOUNTDEVICE}"
-- 
2.32.0

[PATCH 06/10] tests: Change FAT volume label to be with in the valid character range

[Glenn Washburn]

The ';', semi-colon, character is not a valid character for a FAT filesystem
label. This test used to succeed because prior to v4.2 of dosfstools
mkfs.vfat did not enforce the character restrictions for volume labels. So
change the volume label string to be valid but contain symbol characters to
test odd volume labels.

Signed-off-by: Glenn Washburn <development@efficientek.com>
---
 tests/util/grub-fs-tester.in | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/tests/util/grub-fs-tester.in b/tests/util/grub-fs-tester.in
index 96b204590..537cdd48b 100644
--- a/tests/util/grub-fs-tester.in
+++ b/tests/util/grub-fs-tester.in
@@ -343,9 +343,10 @@ for LOGSECSIZE in $(range "$MINLOGSECSIZE" "$MAXLOGSECSIZE" 1); do
 	    # FS LIMITATION: XFS label is at most 12 UTF-8 characters
 		x"xfs"|x"xfs_crc")
 		    FSLABEL="géт 😁к";;
-            # FS LIMITATION: FAT labels limited to 11 characters, no  international characters or lowercase
+            # FS LIMITATION: FAT labels limited to 11 characters, no international characters or lowercase
+	    # and excluding the restricted characters in "*?.,;:/\|+=<>[]"
 		x"vfat"* | xmsdos*)
-		    FSLABEL="GRUBTEST ;_";;
+		    FSLABEL="G~!@#\$%^&(_";;
 	    # FS LIMITATION: AFFS is latin1. At most 29 characters
 		x"affs" | xaffs_intl)
 		    FSLABEL="grub_tt? #*w;/e£@¡¤½¾{[]}<>.,";;
-- 
2.32.0

[PATCH 07/10] tests: Skip hfs test only when mac_roman module is not loaded and not loadable

[Glenn Washburn]

Allow the hfs tests to not be skipped if the mac_roman modules is loaded in
the kernel, but not accessible to modprobe.

Signed-off-by: Glenn Washburn <development@efficientek.com>
---
 tests/hfs_test.in | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/hfs_test.in b/tests/hfs_test.in
index d7ec56bef..5b0c5e33e 100644
--- a/tests/hfs_test.in
+++ b/tests/hfs_test.in
@@ -15,7 +15,7 @@ if ! which mkfs.hfs >/dev/null 2>&1; then
    exit 77
 fi
 
-if ! modprobe mac-roman; then
+if ! grep -q mac_roman /proc/modules && ! modprobe mac_roman; then
    echo "no mac-roman support; cannot test HFS."
    exit 77
 fi
-- 
2.32.0

[PATCH 08/10] grub-shell: Boot PowerPC using PMU instead of CUDA for power management

[Glenn Washburn]

At some point it looks like the defualt machine for qemu-system-ppc started
using CUDA as a backend for power management. This causes the machine to
throw an exception and not actually power down the VM[1]. Switching to PMU
allows power downs and reboots to work as expceted.

[1] https://gitlab.com/qemu-project/qemu/-/issues/624

Signed-off-by: Glenn Washburn <development@efficientek.com>
---
 tests/util/grub-shell.in | 1 +
 1 file changed, 1 insertion(+)

diff --git a/tests/util/grub-shell.in b/tests/util/grub-shell.in
index 93e9f5148..5354d8678 100644
--- a/tests/util/grub-shell.in
+++ b/tests/util/grub-shell.in
@@ -84,6 +84,7 @@ case "${grub_modinfo_target_cpu}-${grub_modinfo_platform}" in
 	serial_null="-serial null"
 	netbootext=elf
 	trim=1
+	qemuopts="-M mac99,via=pmu $qemuopts"
 	;;
 
     sparc64-ieee1275)
-- 
2.32.0

Re: [PATCH 08/10] grub-shell: Boot PowerPC using PMU instead of CUDA for power management

[Paul Menzel]

Dear Glenn,


Thank you. I noticed two small typos.

Am 18.09.21 um 01:04 schrieb Glenn Washburn:
> At some point it looks like the defualt machine for qemu-system-ppc started

default

> using CUDA as a backend for power management. This causes the machine to
> throw an exception and not actually power down the VM[1]. Switching to PMU
> allows power downs and reboots to work as expceted.

expected

> [1] https://gitlab.com/qemu-project/qemu/-/issues/624

Thank you for reporting this upstream.

> Signed-off-by: Glenn Washburn <development@efficientek.com>
> ---
>   tests/util/grub-shell.in | 1 +
>   1 file changed, 1 insertion(+)
> 
> diff --git a/tests/util/grub-shell.in b/tests/util/grub-shell.in
> index 93e9f5148..5354d8678 100644
> --- a/tests/util/grub-shell.in
> +++ b/tests/util/grub-shell.in
> @@ -84,6 +84,7 @@ case "${grub_modinfo_target_cpu}-${grub_modinfo_platform}" in
>   	serial_null="-serial null"
>   	netbootext=elf
>   	trim=1
> +	qemuopts="-M mac99,via=pmu $qemuopts"
>   	;;
>   
>       sparc64-ieee1275)
> 

Reviewed-by: Paul Menzel <pmenzel@molgen.mpg.de>


Kind regards,

Paul

Re: [PATCH 08/10] grub-shell: Boot PowerPC using PMU instead of CUDA for power management

[Glenn Washburn]

On Fri, 17 Sep 2021 23:04:02 +0000
Glenn Washburn <development@efficientek.com> wrote:

> At some point it looks like the defualt machine for qemu-system-ppc started
> using CUDA as a backend for power management. This causes the machine to
> throw an exception and not actually power down the VM[1]. Switching to PMU
> allows power downs and reboots to work as expceted.

As noted in the discussion on the qemu issue, this commit message is
not accurate. So I need to update the message, but the contents of the
patch will remain the same.

Glenn

> 
> [1] https://gitlab.com/qemu-project/qemu/-/issues/624
> 
> Signed-off-by: Glenn Washburn <development@efficientek.com>
> ---
>  tests/util/grub-shell.in | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/tests/util/grub-shell.in b/tests/util/grub-shell.in
> index 93e9f5148..5354d8678 100644
> --- a/tests/util/grub-shell.in
> +++ b/tests/util/grub-shell.in
> @@ -84,6 +84,7 @@ case "${grub_modinfo_target_cpu}-${grub_modinfo_platform}" in
>  	serial_null="-serial null"
>  	netbootext=elf
>  	trim=1
> +	qemuopts="-M mac99,via=pmu $qemuopts"
>  	;;
>  
>      sparc64-ieee1275)

[PATCH 09/10] tests: Output list of devices when partmap fails

[Glenn Washburn]

Signed-off-by: Glenn Washburn <development@efficientek.com>
---
 tests/partmap_test.in | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/partmap_test.in b/tests/partmap_test.in
index faace51ec..3d9871a2f 100644
--- a/tests/partmap_test.in
+++ b/tests/partmap_test.in
@@ -43,7 +43,7 @@ check_output () {
     for dsk in $@; do
 	if ! grep "($dsk)" "${outfile}" >/dev/null
 	then
-	    echo "($dsk): disk/partiton not found"
+	    echo "($dsk): disk/partiton not found in: $(cat "${outfile}")"
 	    exit 1
 	fi
     done
-- 
2.32.0

[PATCH 10/10] tests: Do not delete filesystem images on error

[Glenn Washburn]

The filesystem images created for the filesystem test can be useful when
debugging why a filesystem test failed. So keep them around and let the user
clean them up.

Signed-off-by: Glenn Washburn <development@efficientek.com>
---
 tests/util/grub-fs-tester.in | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/tests/util/grub-fs-tester.in b/tests/util/grub-fs-tester.in
index 537cdd48b..62f814dc4 100644
--- a/tests/util/grub-fs-tester.in
+++ b/tests/util/grub-fs-tester.in
@@ -932,9 +932,6 @@ for LOGSECSIZE in $(range "$MINLOGSECSIZE" "$MAXLOGSECSIZE" 1); do
 				sleep 1
 			    done
 			done
-			for i in $(range 0 $((NDEVICES-1)) 1); do
-			    rm "$FSIMAGEP${i}.img"
-			done
 			exit 99;
 		    fi
 		    ;;
-- 
2.32.0

Re: [PATCH 00/10] More test fixes/improvements

[Daniel Kiper]

On Fri, Sep 17, 2021 at 11:03:54PM +0000, Glenn Washburn wrote:
> Here is another patch series of changes needed to get the make check tests
> passing on fairly recent Ubuntu/Debian systems, though most of the fixes
> are for issues on fairly old systems as well (eg. reiser fixes). The FAT
> volume label change is only needed on fairly recent systems because up until
> very recently a fairly old version of dosfstools has been used which didn't
> do invalid character checking.
>
> Glenn
>
> Glenn Washburn (10):
>   tests: Rename variable filtime -> filetime as its meant to be
>   tests: mkreiserfs only supports 4096 block size
>   tests: Disable reiserfs tests for old format because newer kernels do
>     not support them
>   tests: mkfs.btrfs now supports only 4k sector sizes and above
>   tests: Only test minix3 volumes of 1k block size
>   tests: Change FAT volume label to be with in the valid character range
>   tests: Skip hfs test only when mac_roman module is not loaded and not
>     loadable
>   grub-shell: Boot PowerPC using PMU instead of CUDA for power
>     management
>   tests: Output list of devices when partmap fails
>   tests: Do not delete filesystem images on error

Reviewed-by: Daniel Kiper <daniel.kiper@oracle.com> except patch #8.

You do not need to repost whole series. Just send the patches which did
not get my RB for this and earlier series.

Daniel