From: Wei Gao via ltp <ltp@lists.linux.it>
To: Petr Vorel <pvorel@suse.cz>
Cc: ltp@lists.linux.it
Subject: Re: [LTP] [PATCH v2] ioctl_fiemap01: New test for fiemap ioctl()
Date: Fri, 29 Mar 2024 04:28:35 -0400 [thread overview]
Message-ID: <ZgZ7swC75rxF9hFq@wegao.21.28.32> (raw)
In-Reply-To: <20240228170729.GA1618063@pevik>
On Wed, Feb 28, 2024 at 06:07:29PM +0100, Petr Vorel wrote:
> Hi Wei,
>
> > Fixes: #535
>
> > diff --git a/testcases/kernel/syscalls/ioctl/ioctl_fiemap01.c b/testcases/kernel/syscalls/ioctl/ioctl_fiemap01.c
> > new file mode 100644
> > index 000000000..a626bb03c
> > --- /dev/null
> > +++ b/testcases/kernel/syscalls/ioctl/ioctl_fiemap01.c
> > @@ -0,0 +1,116 @@
> > +// SPDX-License-Identifier: GPL-2.0-only
> > +/*
> > + * Copyright (c) 2023 Wei Gao <wegao@suse.com>
> > + */
> > +
> > +/*\
> > + * [Description]
> > + *
> > + * Verify basic fiemap ioctl
> > + *
> nit: missing space and dot at the end.
> > + */
> > +
> > +#include <linux/fs.h>
> > +#include <linux/fiemap.h>
> > +#include <stdlib.h>
> > +
> > +#include "tst_test.h"
> > +
> > +#define TESTFILE "testfile"
> > +#define NUM_EXTENT 2
> > +#define FILE_OFFSET ((rand() % 8 + 2) * getpagesize())
> > +
> > +static char *buf;
> > +
> > +static void print_extens(struct fiemap *fiemap)
> > +{
> > +
> nit: please no space above.
> > + tst_res(TDEBUG, "File extent count: %u", fiemap->fm_mapped_extents);
> nit: please add space here (readability)
> > + for (unsigned int i = 0; i < fiemap->fm_mapped_extents; ++i) {
> > + tst_res(TDEBUG, "Extent %u: Logical offset: %llu, Physical offset: %llu, flags: %u, Length: %llu",
> > + i + 1,
> > + fiemap->fm_extents[i].fe_logical,
> > + fiemap->fm_extents[i].fe_physical,
> > + fiemap->fm_extents[i].fe_flags,
> > + fiemap->fm_extents[i].fe_length);
> > + }
> > +}
> > +
> > +static void verify_ioctl(void)
> > +{
> > + int fd;
> > + struct fiemap *fiemap;
> > +
> > + fd = SAFE_OPEN(TESTFILE, O_RDWR | O_CREAT, 0644);
> > +
> > + fiemap = SAFE_MALLOC(sizeof(struct fiemap) + sizeof(struct fiemap_extent) * NUM_EXTENT);
> > + fiemap->fm_start = 0;
> > + fiemap->fm_length = ~0ULL;
> > + fiemap->fm_extent_count = 1;
> > +
> > + fiemap->fm_flags = -1;
> > + TST_EXP_FAIL(ioctl(fd, FS_IOC_FIEMAP, fiemap), EBADR);
>
> I get on Tumbleweed with 6.8.0-rc4-1.g9b23bf2-default EOPNOTSUPP:
> ioctl_fiemap01.c:52: TFAIL: ioctl(fd, FS_IOC_FIEMAP, fiemap) expected EBADR: EOPNOTSUPP (95)
> What is the missing dependency for FS_IOC_FIEMAP? Or is there wrong fiemap use
> which causes that?
Thanks for your test and i also reproduce this issue.
This is caused by Tumbleweed mount tmpfs on /tmp and tmpfs seems not support fiemap action.
tmpfs on /tmp type tmpfs (rw,nosuid,nodev,size=494452k,nr_inodes=1048576,inode64)
I will sent new patch for cover all supported filesystem.
>
> > +
> > + fiemap->fm_flags = 0;
> > + SAFE_IOCTL(fd, FS_IOC_FIEMAP, fiemap);
> And on the same kernel another EOPNOTSUPP:
>
> ioctl_fiemap01.c:55: TBROK: ioctl(3,((((2U|1U) << (((0+8)+8)+14)) | ((('f')) << (0+8)) | (((11)) << 0) | ((((sizeof(struct fiemap)))) << ((0+8)+8)))),...) failed: EOPNOTSUPP (95)
> > + print_extens(fiemap);
> > + if (fiemap->fm_mapped_extents == 0)
> > + tst_res(TPASS, "Check fiemap iotct zero fm_mapped_extents pass");
> s/iotct/ioctl/
>
> > + else
> > + tst_res(TFAIL, "Check fiemap iotct zero fm_mapped_extents failed");
> s/iotct/ioctl/
> > +
> > + SAFE_WRITE(SAFE_WRITE_ANY, fd, buf, getpagesize());
> > + SAFE_IOCTL(fd, FS_IOC_FIEMAP, fiemap);
> > + print_extens(fiemap);
> Maybe print only on failure?
> > + if ((fiemap->fm_mapped_extents == 1) && (fiemap->fm_extents[0].fe_physical == 0))
> NOTE: brackets are not needed (== has higher precedence than &&), thus:
> if (fiemap->fm_mapped_extents == 1 && fiemap->fm_extents[0].fe_physical == 0)
>
> > + tst_res(TPASS, "Check fiemap iotct one fm_mapped_extents pass");
> s/iotct/ioctl/
> > + else
> > + tst_res(TFAIL, "Check fiemap iotct one fm_mapped_extents failed");
> s/iotct/ioctl/
> > +
> > + fiemap->fm_flags = FIEMAP_FLAG_SYNC;
> > + SAFE_IOCTL(fd, FS_IOC_FIEMAP, fiemap);
> > + print_extens(fiemap);
> > + if ((fiemap->fm_mapped_extents == 1) &&
> nit: again == does not need to be in ()
> > + (fiemap->fm_extents[0].fe_flags == FIEMAP_EXTENT_LAST) &&
> > + (fiemap->fm_extents[0].fe_physical > 0) &&
> > + (fiemap->fm_extents[0].fe_length == (__u64)getpagesize()))
> > + tst_res(TPASS, "Check fiemap iotct FIEMAP_FLAG_SYNC fm_flags pass");
> s/iotct/ioctl/
> > + else
> > + tst_res(TFAIL, "Check fiemap iotct FIEMAP_FLAG_SYNC fm_flags failed");
> s/iotct/ioctl/
> There are 4x checks, I guess instead it would be worth to write what exactly failed.
>
> > +
> > + fiemap->fm_extent_count = NUM_EXTENT;
> > + srand(time(NULL));
> > + SAFE_LSEEK(fd, FILE_OFFSET, SEEK_SET);
> > + SAFE_WRITE(SAFE_WRITE_ALL, fd, buf, getpagesize());
> > + SAFE_IOCTL(fd, FS_IOC_FIEMAP, fiemap);
> > + print_extens(fiemap);
> > + if ((fiemap->fm_mapped_extents == NUM_EXTENT) &&
> nit: If this check was repeated more than twice, I would put it into separate
> function.
> > + (fiemap->fm_extents[NUM_EXTENT - 1].fe_flags == FIEMAP_EXTENT_LAST) &&
> > + (fiemap->fm_extents[NUM_EXTENT - 1].fe_physical > 0) &&
> > + (fiemap->fm_extents[NUM_EXTENT - 1].fe_length == (__u64)getpagesize()))
> > + tst_res(TPASS, "Check fiemap iotct multiple fm_mapped_extents pass");
> > + else
> > + tst_res(TFAIL, "Check fiemap iotct multiple fm_mapped_extents failed");
> > +
> > + free(fiemap);
> > + SAFE_CLOSE(fd);
> > + unlink(TESTFILE);
> > +}
> ...
>
> Kind regards,
> Petr
--
Mailing list info: https://lists.linux.it/listinfo/ltp
next prev parent reply other threads:[~2024-03-29 8:29 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-10-18 5:43 [LTP] [PATCH v1] ioctl_fiemap01: New test for fiemap ioctl() Wei Gao via ltp
2024-01-15 15:23 ` Cyril Hrubis
2024-01-18 7:32 ` [LTP] [PATCH v2] " Wei Gao via ltp
2024-02-28 17:07 ` Petr Vorel
2024-03-29 8:28 ` Wei Gao via ltp [this message]
2024-03-29 21:32 ` Petr Vorel
2024-03-31 2:15 ` Wei Gao via ltp
2024-03-31 2:17 ` [LTP] [PATCH v3] " Wei Gao via ltp
2024-04-03 9:28 ` Petr Vorel
2024-04-15 10:17 ` Wei Gao via ltp
2024-04-15 12:17 ` Petr Vorel
2024-04-15 11:46 ` [LTP] [PATCH v4] " Wei Gao via ltp
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=ZgZ7swC75rxF9hFq@wegao.21.28.32 \
--to=ltp@lists.linux.it \
--cc=pvorel@suse.cz \
--cc=wegao@suse.com \
/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 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.