[PATCH] libio: Synthesize ESPIPE error if lseek returns 0 after reading bytes

[Frederic Berat]

On Tue, Apr 8, 2025 at 6:38 PM Florian Weimer <fweimer@redhat.com> wrote:

> This is required so that fclose, when trying to seek to the right
> position after filling the input buffer, does not fail with EINVAL.
> This fclose code path only ignores ESPIPE errors.
>
> Reported by Petr Pisar on
> <https://bugzilla.redhat.com/show_bug.cgi?id=2358265>.
>
> Fixes commit be6818be31e756398e45f70e2819d78be0961223 ("Make fclose
> seek input file to right offset (bug 12724)").
>

I executed the test with and without the fix, the behavior matches
expectation.

LGTM.

Reviewed-by: Frédéric Bérat <fberat@redhat.com>


>
> ---
>  libio/fileops.c                   | 10 ++++++++
>  stdio-common/Makefile             |  1 +
>  stdio-common/tst-fclose-devzero.c | 51
> +++++++++++++++++++++++++++++++++++++++
>  3 files changed, 62 insertions(+)
>
> diff --git a/libio/fileops.c b/libio/fileops.c
> index a59e248142..0cce828859 100644
> --- a/libio/fileops.c
> +++ b/libio/fileops.c
> @@ -928,6 +928,16 @@ do_ftell (FILE *fp)
>    if (result == EOF)
>      return result;
>
> +  if (result == 0 && offset < 0)
>

I had to spend some thought on this, but after further discussion it looks
reasonable to me.


> +    {
> +      /* This happens for some character devices that always report
> +        file offset 0 even after some data has been read (instead of
> +        failing with ESPIPE).  The fclose path ignores this
> +        error.  */
> +      __set_errno (ESPIPE);
> +      return EOF;
> +    }
> +
>    result += offset;
>
>    if (result < 0)
> diff --git a/stdio-common/Makefile b/stdio-common/Makefile
> index 3709222266..64b3575acb 100644
> --- a/stdio-common/Makefile
> +++ b/stdio-common/Makefile
> @@ -262,6 +262,7 @@ tests := \
>    tst-bz11319-fortify2 \
>    tst-cookie \
>    tst-dprintf-length \
> +  tst-fclose-devzero \
>    tst-fclose-offset \
>    tst-fdopen \
>    tst-fdopen2 \
> diff --git a/stdio-common/tst-fclose-devzero.c
> b/stdio-common/tst-fclose-devzero.c
> new file mode 100644
> index 0000000000..2a7d808239
> --- /dev/null
> +++ b/stdio-common/tst-fclose-devzero.c
> @@ -0,0 +1,51 @@
> +/* Test that always-zero lseek does not cause fclose failure after fread.
> +   Copyright (C) 2025 Free Software Foundation, Inc.
> +   This file is part of the GNU C Library.
> +
> +   The GNU C Library is free software; you can redistribute it and/or
> +   modify it under the terms of the GNU Lesser General Public
> +   License as published by the Free Software Foundation; either
> +   version 2.1 of the License, or (at your option) any later version.
> +
> +   The GNU C Library is distributed in the hope that it will be useful,
> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> +   Lesser General Public License for more details.
> +
> +   You should have received a copy of the GNU Lesser General Public
> +   License along with the GNU C Library; if not, see
> +   <https://www.gnu.org/licenses/>.  */
> +
> +#include <errno.h>
> +#include <stdio.h>
> +#include <string.h>
> +
> +#include <support/check.h>
> +#include <support/xstdio.h>
> +#include <support/xunistd.h>
> +
> +int
> +do_test (void)
> +{
> +  for (int do_ftello = 0; do_ftello < 2; ++do_ftello)
> +    {
> +      FILE *fp = xfopen ("/dev/zero", "r");
> +      char buf[17];
> +      memset (buf, 0xcc, sizeof (buf));
> +      xfread (buf, 1, sizeof (buf), fp);
> +      static const char zeros[sizeof (buf)] = { 0 };
> +      TEST_COMPARE_BLOB (buf, sizeof (buf), zeros, sizeof (zeros));
> +      if (do_ftello)
> +        {
> +          errno = 0;
> +          TEST_COMPARE (ftello (fp), -1);
> +          TEST_COMPARE (errno, ESPIPE);
> +        }
> +      /* Do not use xfclose because it flushes first.  */
> +      TEST_COMPARE (fclose (fp), 0);
> +    }
> +
> +  return 0;
> +}
> +
> +#include <support/test-driver.c>
>
> base-commit: 4fa959d13d21b8f56a43aa0a416100303736c55c
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://sourceware.org/pipermail/libc-alpha/attachments/20250414/9513fdd0/attachment.htm>