Re: [PATCH v3 2/4] tests: Do not occlude subshell error codes

[Glenn Washburn]

On Mon, 11 Oct 2021 16:20:46 +0200
Daniel Kiper <daniel.kiper@oracle.com> wrote:

> On Thu, Oct 07, 2021 at 03:33:25PM -0500, Glenn Washburn wrote:
> > When a subshell's output is used as input to a "simple command", its return
> > code is not checked. These subshells contain an execution of the grub-shell
> > script which does the work of the actual test. If grub-shell returns an

This "grub-shell" should be "subshell"

> > error code, the test should fail. So refactor to not have the subshell which
> > contains grub-shell be direct input into a simple command (usually the test

This "grub-shell" is correct, as far as I can tell.

> > command). Mostly this is accomplished by having the output first go to a
> > shell variable, and then using the shell variable in the simple command.
> >
> > Signed-off-by: Glenn Washburn <development@efficientek.com>
> > ---
> >  tests/ahci_test.in          | 7 ++++++-
> >  tests/cdboot_test.in        | 3 ++-
> >  tests/core_compress_test.in | 6 ++++--
> >  tests/ehci_test.in          | 7 ++++++-
> >  tests/fddboot_test.in       | 3 ++-
> >  tests/gzcompress_test.in    | 3 ++-
> >  tests/hddboot_test.in       | 3 ++-
> >  tests/lzocompress_test.in   | 3 ++-
> >  tests/netboot_test.in       | 3 ++-
> >  tests/ohci_test.in          | 7 ++++++-
> >  tests/pata_test.in          | 3 ++-
> >  tests/uhci_test.in          | 7 ++++++-
> >  tests/xzcompress_test.in    | 3 ++-
> >  13 files changed, 44 insertions(+), 14 deletions(-)
> >
> > diff --git a/tests/ahci_test.in b/tests/ahci_test.in
> > index 9b1d85df4..1e4e3e443 100644
> > --- a/tests/ahci_test.in
> > +++ b/tests/ahci_test.in
> > @@ -41,7 +41,12 @@ echo "hello" > "$outfile"
> >
> >  tar cf "$imgfile" "$outfile"
> >
> > -if [ "$(echo "nativedisk; source '(ahci0)/$outfile';" | "${grubshell}" 
> > --qemu-opts="-drive id=disk,file=$imgfile,if=none -device ahci,id=ahci 
> > -device ide-hd,drive=disk,bus=ahci.0 " | tail -n 1)" != "Hello World" ]; 
> > then
> > +v=$(echo "nativedisk; source '(ahci0)/$outfile';" |
> > +    "${grubshell}" --qemu-opts="-drive id=disk,file=$imgfile,if=none
> > +                           -device ahci,id=ahci
> > +                           -device ide-hd,drive=disk,bus=ahci.0" |
> > +    tail -n 1)
> 
> The first change partially contradicts what you said in the commit
> message. I had to take a look at the patch #3 to understand why you did
> not drop "tail -n 1". I think the commit message should be clarified in
> a such way which does not require referring to another patch to
> understand this one.

Yep, mix up. Please see correction above to see if it removes the
contradiction you're seeing. The intent was not to refer to a further
patch.

> > +if [ "$v" != "Hello World" ]; then
> >     rm "$imgfile"
> >     rm "$outfile"
> >     exit 1
> > diff --git a/tests/cdboot_test.in b/tests/cdboot_test.in
> > index 75acdfedb..7229f79fb 100644
> > --- a/tests/cdboot_test.in
> > +++ b/tests/cdboot_test.in
> > @@ -34,6 +34,7 @@ case 
> > "${grub_modinfo_target_cpu}-${grub_modinfo_platform}" in
> >     exit 0;;
> >  esac
> >
> > -if [ "$(echo hello | "${grubshell}" --boot=cd)" != "Hello World" ]; then
> > +v=`echo hello | "${grubshell}" --boot=cd`
> 
> I would prefer if you are more consistent and use "$(...)" instead of "`...`".
> The former looks more common in these scripts.

Yep, I agree and prefer in general the former style. Not sure, why I
changed style there. I see its done in a few places. Want me to just
resend this patch or the whole series?

Glenn