* [OE-core][Patch] sstate.bbclass: Check file ownership before doing 'touch -a'
@ 2020-09-25 12:19 Norman
2020-09-25 12:26 ` Andrey Zhizhikin
0 siblings, 1 reply; 6+ messages in thread
From: Norman @ 2020-09-25 12:19 UTC (permalink / raw)
To: openembedded-core
From cd51db253c1a81e47532f1afacce385e422d9a73 Mon Sep 17 00:00:00 2001
From: Norman Stetter <norman.stetter@garz-fricke.com>
Date: Fri, 25 Sep 2020 13:49:43 +0200
Subject: sstate.bbclass: Check file ownership before doing 'touch -a'
In contrast to 'touching' a file without parameters 'touch -a' can only be
performed by the file owner. In case of a shared sstate-cache served
locally (e.g. over NFS), where the user running bitbake is not the
owner of sstate-cache files, even if he has write access on group level,
the sstate_unpack_package task will fail.
Checking if the file is owned by the user before attempting to run
'touch -a' on it solves this.
Signed-off-by: Norman Stetter <norman.stetter@garz-fricke.com>
---
meta/classes/sstate.bbclass | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/meta/classes/sstate.bbclass b/meta/classes/sstate.bbclass
index 375196ef21..66a96a7603 100644
--- a/meta/classes/sstate.bbclass
+++ b/meta/classes/sstate.bbclass
@@ -847,7 +847,7 @@ python sstate_report_unihash() {
sstate_unpack_package () {
tar -xvzf ${SSTATE_PKG}
# update .siginfo atime on local/NFS mirror
- [ -w ${SSTATE_PKG}.siginfo ] && [ -h ${SSTATE_PKG}.siginfo ] && touch -a ${SSTATE_PKG}.siginfo
+ [ -O ${SSTATE_PKG}.siginfo ] && [ -w ${SSTATE_PKG}.siginfo ] && [ -h ${SSTATE_PKG}.siginfo ] && touch -a ${SSTATE_PKG}.siginfo
# Use "! -w ||" to return true for read only files
[ ! -w ${SSTATE_PKG} ] || touch --no-dereference ${SSTATE_PKG}
[ ! -w ${SSTATE_PKG}.sig ] || [ ! -e ${SSTATE_PKG}.sig ] || touch --no-dereference ${SSTATE_PKG}.sig
--
2.17.1
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [OE-core][Patch] sstate.bbclass: Check file ownership before doing 'touch -a'
2020-09-25 12:19 [OE-core][Patch] sstate.bbclass: Check file ownership before doing 'touch -a' Norman
@ 2020-09-25 12:26 ` Andrey Zhizhikin
2020-09-25 12:33 ` Norman
[not found] ` <163806DA121A433C.4559@lists.openembedded.org>
0 siblings, 2 replies; 6+ messages in thread
From: Andrey Zhizhikin @ 2020-09-25 12:26 UTC (permalink / raw)
To: Norman; +Cc: openembedded-core
On Fri, Sep 25, 2020 at 2:19 PM Norman <norman.stetter@garz-fricke.com> wrote:
>
> From cd51db253c1a81e47532f1afacce385e422d9a73 Mon Sep 17 00:00:00 2001
> From: Norman Stetter <norman.stetter@garz-fricke.com>
> Date: Fri, 25 Sep 2020 13:49:43 +0200
> Subject: sstate.bbclass: Check file ownership before doing 'touch -a'
>
> In contrast to 'touching' a file without parameters 'touch -a' can only be
> performed by the file owner. In case of a shared sstate-cache served
> locally (e.g. over NFS), where the user running bitbake is not the
> owner of sstate-cache files, even if he has write access on group level,
> the sstate_unpack_package task will fail.
Isn't the whole intention of using SSTATE_MIRROR to cover exactly
those cases? I do not think that reusing sstate between different
builders via NFS is a proper way to go.
> Checking if the file is owned by the user before attempting to run
> 'touch -a' on it solves this.
>
> Signed-off-by: Norman Stetter <norman.stetter@garz-fricke.com>
> ---
> meta/classes/sstate.bbclass | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/meta/classes/sstate.bbclass b/meta/classes/sstate.bbclass
> index 375196ef21..66a96a7603 100644
> --- a/meta/classes/sstate.bbclass
> +++ b/meta/classes/sstate.bbclass
> @@ -847,7 +847,7 @@ python sstate_report_unihash() {
> sstate_unpack_package () {
> tar -xvzf ${SSTATE_PKG}
> # update .siginfo atime on local/NFS mirror
> - [ -w ${SSTATE_PKG}.siginfo ] && [ -h ${SSTATE_PKG}.siginfo ] && touch -a ${SSTATE_PKG}.siginfo
> + [ -O ${SSTATE_PKG}.siginfo ] && [ -w ${SSTATE_PKG}.siginfo ] && [ -h ${SSTATE_PKG}.siginfo ] && touch -a ${SSTATE_PKG}.siginfo
> # Use "! -w ||" to return true for read only files
> [ ! -w ${SSTATE_PKG} ] || touch --no-dereference ${SSTATE_PKG}
> [ ! -w ${SSTATE_PKG}.sig ] || [ ! -e ${SSTATE_PKG}.sig ] || touch --no-dereference ${SSTATE_PKG}.sig
> --
> 2.17.1
>
>
>
>
--
Regards,
Andrey.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [OE-core][Patch] sstate.bbclass: Check file ownership before doing 'touch -a'
2020-09-25 12:26 ` Andrey Zhizhikin
@ 2020-09-25 12:33 ` Norman
[not found] ` <163806DA121A433C.4559@lists.openembedded.org>
1 sibling, 0 replies; 6+ messages in thread
From: Norman @ 2020-09-25 12:33 UTC (permalink / raw)
To: Andrey Zhizhikin; +Cc: openembedded-core
We do use SSTATE_MIRROR, pointing to a locally mounted drive that gets populated by CI-builds:
SSTATE_MIRRORS = "file://.* file:///var/cache/yocto/sstate-cache/PATH"
> -----Ursprüngliche Nachricht-----
> Von: Andrey Zhizhikin [mailto:andrey.z@gmail.com]
> Gesendet: Friday, September 25, 2020 2:27 PM
> An: Norman Stetter <Norman.Stetter@garz-fricke.com>
> Cc: openembedded-core@lists.openembedded.org
> Betreff: Re: [OE-core][Patch] sstate.bbclass: Check file ownership before
> doing 'touch -a'
>
> On Fri, Sep 25, 2020 at 2:19 PM Norman <norman.stetter@garz-fricke.com>
> wrote:
> >
> > From cd51db253c1a81e47532f1afacce385e422d9a73 Mon Sep 17 00:00:00
> 2001
> > From: Norman Stetter <norman.stetter@garz-fricke.com>
> > Date: Fri, 25 Sep 2020 13:49:43 +0200
> > Subject: sstate.bbclass: Check file ownership before doing 'touch -a'
> >
> > In contrast to 'touching' a file without parameters 'touch -a' can
> > only be performed by the file owner. In case of a shared sstate-cache
> > served locally (e.g. over NFS), where the user running bitbake is not
> > the owner of sstate-cache files, even if he has write access on group
> > level, the sstate_unpack_package task will fail.
>
> Isn't the whole intention of using SSTATE_MIRROR to cover exactly those
> cases? I do not think that reusing sstate between different builders via NFS is
> a proper way to go.
>
> > Checking if the file is owned by the user before attempting to run
> > 'touch -a' on it solves this.
> >
> > Signed-off-by: Norman Stetter <norman.stetter@garz-fricke.com>
> > ---
> > meta/classes/sstate.bbclass | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/meta/classes/sstate.bbclass b/meta/classes/sstate.bbclass
> > index 375196ef21..66a96a7603 100644
> > --- a/meta/classes/sstate.bbclass
> > +++ b/meta/classes/sstate.bbclass
> > @@ -847,7 +847,7 @@ python sstate_report_unihash() {
> > sstate_unpack_package () {
> > tar -xvzf ${SSTATE_PKG}
> > # update .siginfo atime on local/NFS mirror
> > - [ -w ${SSTATE_PKG}.siginfo ] && [ -h ${SSTATE_PKG}.siginfo ] && touch
> -a ${SSTATE_PKG}.siginfo
> > + [ -O ${SSTATE_PKG}.siginfo ] && [ -w ${SSTATE_PKG}.siginfo ]
> > + && [ -h ${SSTATE_PKG}.siginfo ] && touch -a ${SSTATE_PKG}.siginfo
> > # Use "! -w ||" to return true for read only files
> > [ ! -w ${SSTATE_PKG} ] || touch --no-dereference ${SSTATE_PKG}
> > [ ! -w ${SSTATE_PKG}.sig ] || [ ! -e ${SSTATE_PKG}.sig ] ||
> > touch --no-dereference ${SSTATE_PKG}.sig
> > --
> > 2.17.1
> >
> >
> >
> >
>
>
> --
> Regards,
> Andrey.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [OE-core][Patch] sstate.bbclass: Check file ownership before doing 'touch -a'
[not found] ` <163806DA121A433C.4559@lists.openembedded.org>
@ 2020-09-29 6:55 ` Norman
[not found] ` <16392EB84BD69824.17908@lists.openembedded.org>
1 sibling, 0 replies; 6+ messages in thread
From: Norman @ 2020-09-29 6:55 UTC (permalink / raw)
To: Andrey Zhizhikin; +Cc: openembedded-core
I went through it again:
The SSTATE_DIR is set to my local build dir inside my home dir. All files here are owned by me.
The SSTATE_MIRRORS is set to a locally mounted shared sstate-cache. Files here are owned by my group, but not my user.
All *.siginfo files in the SSTATE_DIR are symlinks to the respective files in the SSTATE_MIRRORS dir.
The sstate_unpack_package function gets run for every *.siginfo file in SSTATE_DIR.
If the ${SSTATE_PKG}.siginfo file already exists and is a symlink it's atime is updated (line 850), this affects the destination file of the symlink in SSTATE_MIRRORS, since the --no-dereference switch is not used. The comment in line 849 confirms this is meant to set atime for "local/NFS mirror". But it being a mirror one cannot expect the user running bitbake to be the file owner.
So in my opinion the proposed patch makes sense.
> -----Ursprüngliche Nachricht-----
> Von: openembedded-core@lists.openembedded.org
> [mailto:openembedded-core@lists.openembedded.org] Im Auftrag von
> Norman
> Gesendet: Friday, September 25, 2020 2:33 PM
> An: Andrey Zhizhikin <andrey.z@gmail.com>
> Cc: openembedded-core@lists.openembedded.org
> Betreff: Re: [OE-core][Patch] sstate.bbclass: Check file ownership before
> doing 'touch -a'
>
> We do use SSTATE_MIRROR, pointing to a locally mounted drive that gets
> populated by CI-builds:
>
> SSTATE_MIRRORS = "file://.* file:///var/cache/yocto/sstate-cache/PATH"
>
>
>
> > -----Ursprüngliche Nachricht-----
> > Von: Andrey Zhizhikin [mailto:andrey.z@gmail.com]
> > Gesendet: Friday, September 25, 2020 2:27 PM
> > An: Norman Stetter <Norman.Stetter@garz-fricke.com>
> > Cc: openembedded-core@lists.openembedded.org
> > Betreff: Re: [OE-core][Patch] sstate.bbclass: Check file ownership
> > before doing 'touch -a'
> >
> > On Fri, Sep 25, 2020 at 2:19 PM Norman
> > <norman.stetter@garz-fricke.com>
> > wrote:
> > >
> > > From cd51db253c1a81e47532f1afacce385e422d9a73 Mon Sep 17 00:00:00
> > 2001
> > > From: Norman Stetter <norman.stetter@garz-fricke.com>
> > > Date: Fri, 25 Sep 2020 13:49:43 +0200
> > > Subject: sstate.bbclass: Check file ownership before doing 'touch -a'
> > >
> > > In contrast to 'touching' a file without parameters 'touch -a' can
> > > only be performed by the file owner. In case of a shared
> > > sstate-cache served locally (e.g. over NFS), where the user running
> > > bitbake is not the owner of sstate-cache files, even if he has write
> > > access on group level, the sstate_unpack_package task will fail.
> >
> > Isn't the whole intention of using SSTATE_MIRROR to cover exactly
> > those cases? I do not think that reusing sstate between different
> > builders via NFS is a proper way to go.
> >
> > > Checking if the file is owned by the user before attempting to run
> > > 'touch -a' on it solves this.
> > >
> > > Signed-off-by: Norman Stetter <norman.stetter@garz-fricke.com>
> > > ---
> > > meta/classes/sstate.bbclass | 2 +-
> > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > >
> > > diff --git a/meta/classes/sstate.bbclass
> > > b/meta/classes/sstate.bbclass index 375196ef21..66a96a7603 100644
> > > --- a/meta/classes/sstate.bbclass
> > > +++ b/meta/classes/sstate.bbclass
> > > @@ -847,7 +847,7 @@ python sstate_report_unihash() {
> > > sstate_unpack_package () {
> > > tar -xvzf ${SSTATE_PKG}
> > > # update .siginfo atime on local/NFS mirror
> > > - [ -w ${SSTATE_PKG}.siginfo ] && [ -h ${SSTATE_PKG}.siginfo ] &&
> touch
> > -a ${SSTATE_PKG}.siginfo
> > > + [ -O ${SSTATE_PKG}.siginfo ] && [ -w ${SSTATE_PKG}.siginfo ]
> > > + && [ -h ${SSTATE_PKG}.siginfo ] && touch -a ${SSTATE_PKG}.siginfo
> > > # Use "! -w ||" to return true for read only files
> > > [ ! -w ${SSTATE_PKG} ] || touch --no-dereference ${SSTATE_PKG}
> > > [ ! -w ${SSTATE_PKG}.sig ] || [ ! -e ${SSTATE_PKG}.sig ] ||
> > > touch --no-dereference ${SSTATE_PKG}.sig
> > > --
> > > 2.17.1
> > >
> > >
> > >
> > >
> >
> >
> > --
> > Regards,
> > Andrey.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [OE-core][Patch] sstate.bbclass: Check file ownership before doing 'touch -a'
[not found] ` <16392EB84BD69824.17908@lists.openembedded.org>
@ 2020-09-30 7:17 ` Norman
2020-09-30 10:49 ` Martin Jansa
0 siblings, 1 reply; 6+ messages in thread
From: Norman @ 2020-09-30 7:17 UTC (permalink / raw)
To: Andrey Zhizhikin; +Cc: openembedded-core
I would be happy to get some feedback on this patch.
Cheers,
Norman
> -----Ursprüngliche Nachricht-----
> Von: openembedded-core@lists.openembedded.org
> [mailto:openembedded-core@lists.openembedded.org] Im Auftrag von
> Norman
> Gesendet: Tuesday, September 29, 2020 8:55 AM
> An: Andrey Zhizhikin <andrey.z@gmail.com>
> Cc: openembedded-core@lists.openembedded.org
> Betreff: Re: [OE-core][Patch] sstate.bbclass: Check file ownership before
> doing 'touch -a'
>
> I went through it again:
>
> The SSTATE_DIR is set to my local build dir inside my home dir. All files here
> are owned by me.
> The SSTATE_MIRRORS is set to a locally mounted shared sstate-cache. Files
> here are owned by my group, but not my user.
>
> All *.siginfo files in the SSTATE_DIR are symlinks to the respective files in the
> SSTATE_MIRRORS dir.
>
> The sstate_unpack_package function gets run for every *.siginfo file in
> SSTATE_DIR.
> If the ${SSTATE_PKG}.siginfo file already exists and is a symlink it's atime is
> updated (line 850), this affects the destination file of the symlink in
> SSTATE_MIRRORS, since the --no-dereference switch is not used. The
> comment in line 849 confirms this is meant to set atime for "local/NFS
> mirror". But it being a mirror one cannot expect the user running bitbake to
> be the file owner.
>
> So in my opinion the proposed patch makes sense.
>
>
> > -----Ursprüngliche Nachricht-----
> > Von: openembedded-core@lists.openembedded.org
> > [mailto:openembedded-core@lists.openembedded.org] Im Auftrag von
> > Norman
> > Gesendet: Friday, September 25, 2020 2:33 PM
> > An: Andrey Zhizhikin <andrey.z@gmail.com>
> > Cc: openembedded-core@lists.openembedded.org
> > Betreff: Re: [OE-core][Patch] sstate.bbclass: Check file ownership
> > before doing 'touch -a'
> >
> > We do use SSTATE_MIRROR, pointing to a locally mounted drive that gets
> > populated by CI-builds:
> >
> > SSTATE_MIRRORS = "file://.* file:///var/cache/yocto/sstate-cache/PATH"
> >
> >
> >
> > > -----Ursprüngliche Nachricht-----
> > > Von: Andrey Zhizhikin [mailto:andrey.z@gmail.com]
> > > Gesendet: Friday, September 25, 2020 2:27 PM
> > > An: Norman Stetter <Norman.Stetter@garz-fricke.com>
> > > Cc: openembedded-core@lists.openembedded.org
> > > Betreff: Re: [OE-core][Patch] sstate.bbclass: Check file ownership
> > > before doing 'touch -a'
> > >
> > > On Fri, Sep 25, 2020 at 2:19 PM Norman
> > > <norman.stetter@garz-fricke.com>
> > > wrote:
> > > >
> > > > From cd51db253c1a81e47532f1afacce385e422d9a73 Mon Sep 17
> 00:00:00
> > > 2001
> > > > From: Norman Stetter <norman.stetter@garz-fricke.com>
> > > > Date: Fri, 25 Sep 2020 13:49:43 +0200
> > > > Subject: sstate.bbclass: Check file ownership before doing 'touch -a'
> > > >
> > > > In contrast to 'touching' a file without parameters 'touch -a' can
> > > > only be performed by the file owner. In case of a shared
> > > > sstate-cache served locally (e.g. over NFS), where the user
> > > > running bitbake is not the owner of sstate-cache files, even if he
> > > > has write access on group level, the sstate_unpack_package task will
> fail.
> > >
> > > Isn't the whole intention of using SSTATE_MIRROR to cover exactly
> > > those cases? I do not think that reusing sstate between different
> > > builders via NFS is a proper way to go.
> > >
> > > > Checking if the file is owned by the user before attempting to run
> > > > 'touch -a' on it solves this.
> > > >
> > > > Signed-off-by: Norman Stetter <norman.stetter@garz-fricke.com>
> > > > ---
> > > > meta/classes/sstate.bbclass | 2 +-
> > > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > > >
> > > > diff --git a/meta/classes/sstate.bbclass
> > > > b/meta/classes/sstate.bbclass index 375196ef21..66a96a7603 100644
> > > > --- a/meta/classes/sstate.bbclass
> > > > +++ b/meta/classes/sstate.bbclass
> > > > @@ -847,7 +847,7 @@ python sstate_report_unihash() {
> > > > sstate_unpack_package () {
> > > > tar -xvzf ${SSTATE_PKG}
> > > > # update .siginfo atime on local/NFS mirror
> > > > - [ -w ${SSTATE_PKG}.siginfo ] && [ -h ${SSTATE_PKG}.siginfo ] &&
> > touch
> > > -a ${SSTATE_PKG}.siginfo
> > > > + [ -O ${SSTATE_PKG}.siginfo ] && [ -w ${SSTATE_PKG}.siginfo
> > > > + ] && [ -h ${SSTATE_PKG}.siginfo ] && touch -a
> > > > + ${SSTATE_PKG}.siginfo
> > > > # Use "! -w ||" to return true for read only files
> > > > [ ! -w ${SSTATE_PKG} ] || touch --no-dereference ${SSTATE_PKG}
> > > > [ ! -w ${SSTATE_PKG}.sig ] || [ ! -e ${SSTATE_PKG}.sig ]
> > > > || touch --no-dereference ${SSTATE_PKG}.sig
> > > > --
> > > > 2.17.1
> > > >
> > > >
> > > >
> > > >
> > >
> > >
> > > --
> > > Regards,
> > > Andrey.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [OE-core][Patch] sstate.bbclass: Check file ownership before doing 'touch -a'
2020-09-30 7:17 ` Norman
@ 2020-09-30 10:49 ` Martin Jansa
0 siblings, 0 replies; 6+ messages in thread
From: Martin Jansa @ 2020-09-30 10:49 UTC (permalink / raw)
To: Norman; +Cc: Andrey Zhizhikin, openembedded-core
[-- Attachment #1: Type: text/plain, Size: 5295 bytes --]
It was merged 2 days ago:
https://git.openembedded.org/openembedded-core/commit/?id=f1554e8bd40c640fd58daac18ecf7b458e754109
On Wed, Sep 30, 2020 at 9:17 AM Norman <norman.stetter@garz-fricke.com>
wrote:
> I would be happy to get some feedback on this patch.
>
> Cheers,
>
> Norman
>
> > -----Ursprüngliche Nachricht-----
> > Von: openembedded-core@lists.openembedded.org
> > [mailto:openembedded-core@lists.openembedded.org] Im Auftrag von
> > Norman
> > Gesendet: Tuesday, September 29, 2020 8:55 AM
> > An: Andrey Zhizhikin <andrey.z@gmail.com>
> > Cc: openembedded-core@lists.openembedded.org
> > Betreff: Re: [OE-core][Patch] sstate.bbclass: Check file ownership before
> > doing 'touch -a'
> >
> > I went through it again:
> >
> > The SSTATE_DIR is set to my local build dir inside my home dir. All
> files here
> > are owned by me.
> > The SSTATE_MIRRORS is set to a locally mounted shared sstate-cache. Files
> > here are owned by my group, but not my user.
> >
> > All *.siginfo files in the SSTATE_DIR are symlinks to the respective
> files in the
> > SSTATE_MIRRORS dir.
> >
> > The sstate_unpack_package function gets run for every *.siginfo file in
> > SSTATE_DIR.
> > If the ${SSTATE_PKG}.siginfo file already exists and is a symlink it's
> atime is
> > updated (line 850), this affects the destination file of the symlink in
> > SSTATE_MIRRORS, since the --no-dereference switch is not used. The
> > comment in line 849 confirms this is meant to set atime for "local/NFS
> > mirror". But it being a mirror one cannot expect the user running
> bitbake to
> > be the file owner.
> >
> > So in my opinion the proposed patch makes sense.
> >
> >
> > > -----Ursprüngliche Nachricht-----
> > > Von: openembedded-core@lists.openembedded.org
> > > [mailto:openembedded-core@lists.openembedded.org] Im Auftrag von
> > > Norman
> > > Gesendet: Friday, September 25, 2020 2:33 PM
> > > An: Andrey Zhizhikin <andrey.z@gmail.com>
> > > Cc: openembedded-core@lists.openembedded.org
> > > Betreff: Re: [OE-core][Patch] sstate.bbclass: Check file ownership
> > > before doing 'touch -a'
> > >
> > > We do use SSTATE_MIRROR, pointing to a locally mounted drive that gets
> > > populated by CI-builds:
> > >
> > > SSTATE_MIRRORS = "file://.* file:///var/cache/yocto/sstate-cache/PATH"
> > >
> > >
> > >
> > > > -----Ursprüngliche Nachricht-----
> > > > Von: Andrey Zhizhikin [mailto:andrey.z@gmail.com]
> > > > Gesendet: Friday, September 25, 2020 2:27 PM
> > > > An: Norman Stetter <Norman.Stetter@garz-fricke.com>
> > > > Cc: openembedded-core@lists.openembedded.org
> > > > Betreff: Re: [OE-core][Patch] sstate.bbclass: Check file ownership
> > > > before doing 'touch -a'
> > > >
> > > > On Fri, Sep 25, 2020 at 2:19 PM Norman
> > > > <norman.stetter@garz-fricke.com>
> > > > wrote:
> > > > >
> > > > > From cd51db253c1a81e47532f1afacce385e422d9a73 Mon Sep 17
> > 00:00:00
> > > > 2001
> > > > > From: Norman Stetter <norman.stetter@garz-fricke.com>
> > > > > Date: Fri, 25 Sep 2020 13:49:43 +0200
> > > > > Subject: sstate.bbclass: Check file ownership before doing 'touch
> -a'
> > > > >
> > > > > In contrast to 'touching' a file without parameters 'touch -a' can
> > > > > only be performed by the file owner. In case of a shared
> > > > > sstate-cache served locally (e.g. over NFS), where the user
> > > > > running bitbake is not the owner of sstate-cache files, even if he
> > > > > has write access on group level, the sstate_unpack_package task
> will
> > fail.
> > > >
> > > > Isn't the whole intention of using SSTATE_MIRROR to cover exactly
> > > > those cases? I do not think that reusing sstate between different
> > > > builders via NFS is a proper way to go.
> > > >
> > > > > Checking if the file is owned by the user before attempting to run
> > > > > 'touch -a' on it solves this.
> > > > >
> > > > > Signed-off-by: Norman Stetter <norman.stetter@garz-fricke.com>
> > > > > ---
> > > > > meta/classes/sstate.bbclass | 2 +-
> > > > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > > > >
> > > > > diff --git a/meta/classes/sstate.bbclass
> > > > > b/meta/classes/sstate.bbclass index 375196ef21..66a96a7603 100644
> > > > > --- a/meta/classes/sstate.bbclass
> > > > > +++ b/meta/classes/sstate.bbclass
> > > > > @@ -847,7 +847,7 @@ python sstate_report_unihash() {
> > > > > sstate_unpack_package () {
> > > > > tar -xvzf ${SSTATE_PKG}
> > > > > # update .siginfo atime on local/NFS mirror
> > > > > - [ -w ${SSTATE_PKG}.siginfo ] && [ -h ${SSTATE_PKG}.siginfo
> ] &&
> > > touch
> > > > -a ${SSTATE_PKG}.siginfo
> > > > > + [ -O ${SSTATE_PKG}.siginfo ] && [ -w ${SSTATE_PKG}.siginfo
> > > > > + ] && [ -h ${SSTATE_PKG}.siginfo ] && touch -a
> > > > > + ${SSTATE_PKG}.siginfo
> > > > > # Use "! -w ||" to return true for read only files
> > > > > [ ! -w ${SSTATE_PKG} ] || touch --no-dereference
> ${SSTATE_PKG}
> > > > > [ ! -w ${SSTATE_PKG}.sig ] || [ ! -e ${SSTATE_PKG}.sig ]
> > > > > || touch --no-dereference ${SSTATE_PKG}.sig
> > > > > --
> > > > > 2.17.1
> > > > >
> > > > >
> > > > >
> > > > >
> > > >
> > > >
> > > > --
> > > > Regards,
> > > > Andrey.
>
>
>
>
[-- Attachment #2: Type: text/html, Size: 8043 bytes --]
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2020-09-30 10:49 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-09-25 12:19 [OE-core][Patch] sstate.bbclass: Check file ownership before doing 'touch -a' Norman
2020-09-25 12:26 ` Andrey Zhizhikin
2020-09-25 12:33 ` Norman
[not found] ` <163806DA121A433C.4559@lists.openembedded.org>
2020-09-29 6:55 ` Norman
[not found] ` <16392EB84BD69824.17908@lists.openembedded.org>
2020-09-30 7:17 ` Norman
2020-09-30 10:49 ` Martin Jansa
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.