diff mbox series

fsstress: improve error message on check_cwd() error

Message ID 20211101155559.3988492-1-mcgrof@kernel.org (mailing list archive)
State New, archived
Headers show
Series fsstress: improve error message on check_cwd() error | expand

Commit Message

Luis Chamberlain Nov. 1, 2021, 3:55 p.m. UTC
I ran into an error with generic/083 with xfs due to check_cwd() but
why it failed is not clear because there are two types of
failures:

  o stat64() failed (likely -ENOMEM is my guess)
  o the inode actually changed

Throw a bone out to developers so that in case en error does happen
they know which rabbit hole to go down on.

Cc: Anthony Iliopoulos <ailiopoulos@suse.de>
Signed-off-by: Luis Chamberlain <mcgrof@kernel.org>
---
 ltp/fsstress.c | 15 ++++++++++++++-
 1 file changed, 14 insertions(+), 1 deletion(-)

Comments

Darrick J. Wong Nov. 3, 2021, 4:24 p.m. UTC | #1
On Mon, Nov 01, 2021 at 08:55:59AM -0700, Luis Chamberlain wrote:
> I ran into an error with generic/083 with xfs due to check_cwd() but
> why it failed is not clear because there are two types of
> failures:
> 
>   o stat64() failed (likely -ENOMEM is my guess)
>   o the inode actually changed
> 
> Throw a bone out to developers so that in case en error does happen
> they know which rabbit hole to go down on.

<cough> word choice on those last three words...

> 
> Cc: Anthony Iliopoulos <ailiopoulos@suse.de>
> Signed-off-by: Luis Chamberlain <mcgrof@kernel.org>
> ---
>  ltp/fsstress.c | 15 ++++++++++++++-
>  1 file changed, 14 insertions(+), 1 deletion(-)
> 
> diff --git a/ltp/fsstress.c b/ltp/fsstress.c
> index 90ae432e..a576afea 100644
> --- a/ltp/fsstress.c
> +++ b/ltp/fsstress.c
> @@ -9,6 +9,7 @@
>  #include <sys/uio.h>
>  #include <stddef.h>
>  #include <stdbool.h>
> +#include <string.h>
>  #include "global.h"
>  
>  #ifdef HAVE_BTRFSUTIL_H
> @@ -943,9 +944,21 @@ check_cwd(void)
>  {
>  #ifdef DEBUG
>  	struct stat64	statbuf;
> +	int ret;
> +
> +	ret = stat64(".", &statbuf);
> +	if (ret !=0) {

Nit: space between '!=' and '0'.

> +		fprintf(stderr, "fsstress: check_cwd stat64 failed with: %d (%s)\n",
> +			ret, strerror(ret));

ret is set to -1 on error, according to the manpage; to get the real
error you'd have to call strerror(errno) as the last arg, or be lazy
and:

		perror("fsstress check_cwd stat64");

> +		goto out;
> +	}
>  
> -	if (stat64(".", &statbuf) == 0 && statbuf.st_ino == top_ino)
> +	if (statbuf.st_ino == top_ino)
>  		return;
> +
> +	fprintf(stderr, "fsstress: check_cwd statbuf.st_ino (%lu) != top_ino (%lu)\n",
> +		statbuf.st_ino, top_ino);

This might want some explicit casting, since this can be defined as
anything between unsigned long to uint64_t, at least according to the
glibc headers on my system.

--D

> +out:
>  	assert(chdir(homedir) == 0);
>  	fprintf(stderr, "fsstress: check_cwd failure\n");
>  	abort();
> -- 
> 2.33.0
>
Luis Chamberlain Nov. 3, 2021, 6:18 p.m. UTC | #2
On Wed, Nov 03, 2021 at 09:24:34AM -0700, Darrick J. Wong wrote:
> On Mon, Nov 01, 2021 at 08:55:59AM -0700, Luis Chamberlain wrote:
> > I ran into an error with generic/083 with xfs due to check_cwd() but
> > why it failed is not clear because there are two types of
> > failures:
> > 
> >   o stat64() failed (likely -ENOMEM is my guess)
> >   o the inode actually changed
> > 
> > Throw a bone out to developers so that in case en error does happen
> > they know which rabbit hole to go down on.
> 
> <cough> word choice on those last three words...

Will fix thanks.

> > 
> > Cc: Anthony Iliopoulos <ailiopoulos@suse.de>
> > Signed-off-by: Luis Chamberlain <mcgrof@kernel.org>
> > ---
> >  ltp/fsstress.c | 15 ++++++++++++++-
> >  1 file changed, 14 insertions(+), 1 deletion(-)
> > 
> > diff --git a/ltp/fsstress.c b/ltp/fsstress.c
> > index 90ae432e..a576afea 100644
> > --- a/ltp/fsstress.c
> > +++ b/ltp/fsstress.c
> > @@ -9,6 +9,7 @@
> >  #include <sys/uio.h>
> >  #include <stddef.h>
> >  #include <stdbool.h>
> > +#include <string.h>
> >  #include "global.h"
> >  
> >  #ifdef HAVE_BTRFSUTIL_H
> > @@ -943,9 +944,21 @@ check_cwd(void)
> >  {
> >  #ifdef DEBUG
> >  	struct stat64	statbuf;
> > +	int ret;
> > +
> > +	ret = stat64(".", &statbuf);
> > +	if (ret !=0) {
> 
> Nit: space between '!=' and '0'.

OK.

> > +		fprintf(stderr, "fsstress: check_cwd stat64 failed with: %d (%s)\n",
> > +			ret, strerror(ret));
> 
> ret is set to -1 on error, according to the manpage; to get the real
> error you'd have to call strerror(errno) as the last arg, or be lazy
> and:

I don't want to be lazy here as this is a real issue.

> 		perror("fsstress check_cwd stat64");
> 
> > +		goto out;
> > +	}
> >  
> > -	if (stat64(".", &statbuf) == 0 && statbuf.st_ino == top_ino)
> > +	if (statbuf.st_ino == top_ino)
> >  		return;
> > +
> > +	fprintf(stderr, "fsstress: check_cwd statbuf.st_ino (%lu) != top_ino (%lu)\n",
> > +		statbuf.st_ino, top_ino);
> 
> This might want some explicit casting, since this can be defined as
> anything between unsigned long to uint64_t, at least according to the
> glibc headers on my system.

Um, Filipe had suggested something a bit different before. Can you guys
decide and let me know your final preference ? :)

  Luis
Luis Chamberlain Nov. 5, 2021, 4:04 p.m. UTC | #3
On Wed, Nov 03, 2021 at 11:18:22AM -0700, Luis Chamberlain wrote:
> On Wed, Nov 03, 2021 at 09:24:34AM -0700, Darrick J. Wong wrote:
> > > -	if (stat64(".", &statbuf) == 0 && statbuf.st_ino == top_ino)
> > > +	if (statbuf.st_ino == top_ino)
> > >  		return;
> > > +
> > > +	fprintf(stderr, "fsstress: check_cwd statbuf.st_ino (%lu) != top_ino (%lu)\n",
> > > +		statbuf.st_ino, top_ino);
> > 
> > This might want some explicit casting, since this can be defined as
> > anything between unsigned long to uint64_t, at least according to the
> > glibc headers on my system.
> 
> Um, Filipe had suggested something a bit different before. Can you guys
> decide and let me know your final preference ? :)

Any final preference / recommendation?

  Luis
Darrick J. Wong Nov. 5, 2021, 4:09 p.m. UTC | #4
On Fri, Nov 05, 2021 at 09:04:21AM -0700, Luis Chamberlain wrote:
> On Wed, Nov 03, 2021 at 11:18:22AM -0700, Luis Chamberlain wrote:
> > On Wed, Nov 03, 2021 at 09:24:34AM -0700, Darrick J. Wong wrote:
> > > > -	if (stat64(".", &statbuf) == 0 && statbuf.st_ino == top_ino)
> > > > +	if (statbuf.st_ino == top_ino)
> > > >  		return;
> > > > +
> > > > +	fprintf(stderr, "fsstress: check_cwd statbuf.st_ino (%lu) != top_ino (%lu)\n",
> > > > +		statbuf.st_ino, top_ino);
> > > 
> > > This might want some explicit casting, since this can be defined as
> > > anything between unsigned long to uint64_t, at least according to the
> > > glibc headers on my system.
> > 
> > Um, Filipe had suggested something a bit different before. Can you guys
> > decide and let me know your final preference ? :)
> 
> Any final preference / recommendation?

Use %llu with explicit casts to (unsigned long long) and it should work
fine everywhere.

--D

>   Luis
diff mbox series

Patch

diff --git a/ltp/fsstress.c b/ltp/fsstress.c
index 90ae432e..a576afea 100644
--- a/ltp/fsstress.c
+++ b/ltp/fsstress.c
@@ -9,6 +9,7 @@ 
 #include <sys/uio.h>
 #include <stddef.h>
 #include <stdbool.h>
+#include <string.h>
 #include "global.h"
 
 #ifdef HAVE_BTRFSUTIL_H
@@ -943,9 +944,21 @@  check_cwd(void)
 {
 #ifdef DEBUG
 	struct stat64	statbuf;
+	int ret;
+
+	ret = stat64(".", &statbuf);
+	if (ret !=0) {
+		fprintf(stderr, "fsstress: check_cwd stat64 failed with: %d (%s)\n",
+			ret, strerror(ret));
+		goto out;
+	}
 
-	if (stat64(".", &statbuf) == 0 && statbuf.st_ino == top_ino)
+	if (statbuf.st_ino == top_ino)
 		return;
+
+	fprintf(stderr, "fsstress: check_cwd statbuf.st_ino (%lu) != top_ino (%lu)\n",
+		statbuf.st_ino, top_ino);
+out:
 	assert(chdir(homedir) == 0);
 	fprintf(stderr, "fsstress: check_cwd failure\n");
 	abort();