From: "Roberto A. Foglietta" <roberto.foglietta@gmail.com>
To: Henning Schild <henning.schild@siemens.com>
Cc: Uladzimir Bely <ubely@ilbers.de>, isar-users@googlegroups.com
Subject: Re: [PATCH] deb-dl-dir: Fallback to copying when import-export
Date: Wed, 1 Feb 2023 15:27:01 +0100 [thread overview]
Message-ID: <CAJGKYO4zcFD+PRpQDeV1WkJLTiPJKZk6J4EosFyQT0BpbVifJQ@mail.gmail.com> (raw)
In-Reply-To: <20230201100046.27c229b6@md1za8fc.ad001.siemens.net>
On Wed, 1 Feb 2023 at 10:00, Henning Schild <henning.schild@siemens.com> wrote:
>
> Am Tue, 31 Jan 2023 12:20:12 +0100
> schrieb Uladzimir Bely <ubely@ilbers.de>:
>
> > By default, hardlinks are used when doing DL_DIR import and export.
> > This fails if DL_DIR's filesystem differs from TMPDIR's one.
> >
> > Fallback to file copying (with lower performance and huge disk usage)
> > in this case.
> >
> > Signed-off-by: Uladzimir Bely <ubely@ilbers.de>
> > ---
> > meta/classes/deb-dl-dir.bbclass | 18 ++++++++++++++++--
> > 1 file changed, 16 insertions(+), 2 deletions(-)
> >
> > diff --git a/meta/classes/deb-dl-dir.bbclass
> > b/meta/classes/deb-dl-dir.bbclass index 7db25251..d572e6fb 100644
> > --- a/meta/classes/deb-dl-dir.bbclass
> > +++ b/meta/classes/deb-dl-dir.bbclass
> > @@ -80,13 +80,20 @@ deb_dl_dir_import() {
> > export pc="${DEBDIR}/${2}"
> > export rootfs="${1}"
> > sudo mkdir -p "${rootfs}"/var/cache/apt/archives/
> > +
> > + # Use hardlinks with fallback to `cp` if different filesystems
> > + export cmd="ln -Pf -t"
> > + if [ "$(stat -c %m ${pc})" != "$(stat -c %m ${rootfs})" ]; then
>
> Instead of dealing only with the known problem and expecting it to
> come, maybe try an actual hardlink and "catch the exception".
>
Hi Henning,
> cmd = ln
cmd = "ln -Pft $dest" but this is not the issue
> cmd testfile || cmd = cp
It might not work because cp -l complains and fails but ln -P
complains but succeeds (or at least succeeds enough even in some cases
in which it complains). IMHO, if it complains it also returns an error
as well. But also this is not the issue.
> rm -f testfile
>
> Who knows which other reasons we might find where hardlinks do not
> work. And who knows how that stat could get confused by bind-mounts or
> funny filesystems ... where hardlinks might work after all, even if the
> stat would guess not.
>
> So instead of guessing i am all for actually trying.
>
The general approach you are suggesting is valid and it is good advice.
Best regards, R-
prev parent reply other threads:[~2023-02-01 14:27 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-01-31 11:20 Uladzimir Bely
2023-02-01 9:00 ` Henning Schild
2023-02-01 14:27 ` Roberto A. Foglietta [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CAJGKYO4zcFD+PRpQDeV1WkJLTiPJKZk6J4EosFyQT0BpbVifJQ@mail.gmail.com \
--to=roberto.foglietta@gmail.com \
--cc=henning.schild@siemens.com \
--cc=isar-users@googlegroups.com \
--cc=ubely@ilbers.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox