Hi Thomas,
This is a pull request for the series that I sent earlier.
The series aims to switch vfs timestamps to use
struct timespec64. Currently vfs uses struct timespec,
which is not y2038 safe.
The flag patch applies cleanly. I've not seen the timestamps
update logic change often. The series applies cleanly on 4.17-rc6
and linux-next tip (top commit: next-20180517).
I'm not sure how to merge this kind of a series with a flag patch.
We are targeting 4.18 for this.
Let me know if you have other suggestions.
The series involves the following:
1. Add vfs helper functions for supporting struct timepec64 timestamps.
2. Cast prints of vfs timestamps to avoid warnings after the switch.
3. Simplify code using vfs timestamps so that the actual
replacement becomes easy.
4. Convert vfs timestamps to use struct timespec64 using a script.
This is a flag day patch.
I've tried to keep the conversions with the script simple, to
aid in the reviews. I've kept all the internal filesystem data
structures and function signatures the same.
Next steps:
1. Convert APIs that can handle timespec64, instead of converting
timestamps at the boundaries.
2. Update internal data structures to avoid timestamp conversions.
--
Changes from v1:
* Fix the pointer bug in the udf patch.
* Include Kees's patch for pstore.
* Fix hostfs regression found by kbuild bot.
The following changes since commit 771c577c23bac90597c685971d7297ea00f99d11:
Linux 4.17-rc6 (2018-05-20 15:31:38 -0700)
are available in the Git repository at:
https://github.com/deepa-hub/vfs vfs_timespec64
for you to fetch changes up to 213ae530f442029284ab6041812df0571ebd9da7:
vfs: change inode times to use struct timespec64 (2018-05-25 15:31:14 -0700)
----------------------------------------------------------------
Deepa Dinamani (6):
fs: add timespec64_truncate()
lustre: Use long long type to print inode time
ceph: make inode time prints to be long long
fs: nfs: get rid of memcpys for inode times
udf: Simplify calls to udf_disk_stamp_to_time
vfs: change inode times to use struct timespec64
Kees Cook (1):
pstore: Convert internal records to timespec64
drivers/firmware/efi/efi-pstore.c | 27 ++++----
drivers/staging/lustre/lustre/llite/llite_lib.c | 12 ++--
drivers/staging/lustre/lustre/llite/namei.c | 5 +-
drivers/staging/lustre/lustre/lmv/lmv_obd.c | 7 +-
drivers/staging/lustre/lustre/mdc/mdc_reint.c | 6 +-
drivers/staging/lustre/lustre/obdclass/obdo.c | 6 +-
drivers/tty/tty_io.c | 15 ++++-
drivers/usb/gadget/function/f_fs.c | 2 +-
fs/adfs/inode.c | 7 +-
fs/afs/fsclient.c | 2 +-
fs/attr.c | 14 ++--
fs/bad_inode.c | 2 +-
fs/btrfs/file.c | 6 +-
fs/btrfs/inode.c | 8 +--
fs/btrfs/ioctl.c | 4 +-
fs/btrfs/root-tree.c | 4 +-
fs/btrfs/transaction.c | 2 +-
fs/ceph/addr.c | 12 ++--
fs/ceph/cache.c | 4 +-
fs/ceph/caps.c | 6 +-
fs/ceph/file.c | 6 +-
fs/ceph/inode.c | 86 +++++++++++++------------
fs/ceph/mds_client.c | 7 +-
fs/ceph/snap.c | 6 +-
fs/cifs/cache.c | 4 +-
fs/cifs/fscache.c | 8 +--
fs/cifs/inode.c | 26 ++++----
fs/coda/coda_linux.c | 12 ++--
fs/configfs/inode.c | 12 ++--
fs/cramfs/inode.c | 2 +-
fs/ext4/ext4.h | 34 ++++++----
fs/ext4/ialloc.c | 4 +-
fs/ext4/namei.c | 2 +-
fs/f2fs/f2fs.h | 10 ++-
fs/f2fs/file.c | 12 ++--
fs/f2fs/inode.c | 12 ++--
fs/f2fs/namei.c | 4 +-
fs/fat/inode.c | 20 ++++--
fs/fat/namei_msdos.c | 21 +++---
fs/fat/namei_vfat.c | 22 ++++---
fs/fuse/inode.c | 2 +-
fs/gfs2/dir.c | 6 +-
fs/gfs2/glops.c | 4 +-
fs/hfs/inode.c | 4 +-
fs/hfsplus/inode.c | 12 ++--
fs/hostfs/hostfs_kern.c | 12 ++--
fs/inode.c | 58 ++++++++++++-----
fs/jffs2/dir.c | 18 +++---
fs/jffs2/file.c | 2 +-
fs/jffs2/fs.c | 12 ++--
fs/kernfs/dir.c | 4 +-
fs/kernfs/inode.c | 8 +--
fs/locks.c | 2 +-
fs/nfs/callback_proc.c | 4 +-
fs/nfs/fscache-index.c | 4 +-
fs/nfs/fscache.c | 12 ++--
fs/nfs/inode.c | 39 ++++++-----
fs/nfs/nfs2xdr.c | 25 ++++---
fs/nfs/nfs3xdr.c | 8 ++-
fs/nfs/nfs4xdr.c | 7 +-
fs/nfsd/blocklayout.c | 8 ++-
fs/nfsd/nfs3xdr.c | 14 ++--
fs/nfsd/nfs4xdr.c | 7 +-
fs/nfsd/nfsxdr.c | 2 +-
fs/ntfs/inode.c | 30 ++++-----
fs/ocfs2/dlmglue.c | 20 ++++--
fs/ocfs2/file.c | 6 +-
fs/orangefs/inode.c | 2 +-
fs/orangefs/orangefs-kernel.h | 2 +-
fs/overlayfs/inode.c | 2 +-
fs/overlayfs/overlayfs.h | 2 +-
fs/proc/uptime.c | 2 +-
fs/pstore/platform.c | 2 +-
fs/pstore/ram.c | 21 ++++--
fs/reiserfs/namei.c | 2 +-
fs/reiserfs/xattr.c | 4 +-
fs/ubifs/dir.c | 4 +-
fs/ubifs/file.c | 23 +++----
fs/ubifs/ubifs.h | 2 +-
fs/udf/ialloc.c | 4 +-
fs/udf/inode.c | 59 +++++++++--------
fs/udf/super.c | 17 +++--
fs/udf/udfdecl.h | 4 +-
fs/udf/udftime.c | 9 +--
fs/xfs/xfs_inode.c | 2 +-
fs/xfs/xfs_iops.c | 2 +-
fs/xfs/xfs_trans_inode.c | 2 +-
include/linux/fs.h | 24 +++----
include/linux/pstore.h | 2 +-
include/linux/stat.h | 8 +--
90 files changed, 558 insertions(+), 434 deletions(-)
--
2.17.0
The pstore conversion to timespec64 introduces its own method of passing
seconds into sscanf() and sprintf() type functions to work around the
timespec64 definition on 64-bit systems that redefine it to 'timespec'.
That hack is now finally getting removed, but that means we get a (harmless)
warning once both patches are merged:
fs/pstore/ram.c: In function 'ramoops_read_kmsg_hdr':
fs/pstore/ram.c:39:29: error: format '%ld' expects argument of type 'long int *', but argument 3 has type 'time64_t *' {aka 'long long int *'} [-Werror=format=]
#define RAMOOPS_KERNMSG_HDR "===="
^~~~~~
fs/pstore/ram.c:167:21: note: in expansion of macro 'RAMOOPS_KERNMSG_HDR'
This removes the pstore specific workaround and uses the same method that
we have in place for all other functions that print a timespec64.
Related to this, I found that the kasprintf() output contains an incorrect
nanosecond values for any number starting with zeroes, and I adapt the
format string accordingly.
Link: https://lkml.org/lkml/2018/5/19/115
Link: https://lkml.org/lkml/2018/5/16/1080
Fixes: 0f0d83b99ef7 ("pstore: Convert internal records to timespec64")
Signed-off-by: Arnd Bergmann <arnd(a)arndb.de>
---
fs/pstore/ram.c | 17 ++++++-----------
1 file changed, 6 insertions(+), 11 deletions(-)
diff --git a/fs/pstore/ram.c b/fs/pstore/ram.c
index 69e893076ab7..bbd1e357c23d 100644
--- a/fs/pstore/ram.c
+++ b/fs/pstore/ram.c
@@ -38,11 +38,6 @@
#define RAMOOPS_KERNMSG_HDR "===="
#define MIN_MEM_SIZE 4096UL
-#if __BITS_PER_LONG == 64
-# define TVSEC_FMT "%ld"
-#else
-# define TVSEC_FMT "%lld"
-#endif
static ulong record_size = MIN_MEM_SIZE;
module_param(record_size, ulong, 0400);
@@ -164,15 +159,15 @@ static int ramoops_read_kmsg_hdr(char *buffer, struct timespec64 *time,
char data_type;
int header_length = 0;
- if (sscanf(buffer, RAMOOPS_KERNMSG_HDR TVSEC_FMT ".%lu-%c\n%n",
- &time->tv_sec, &time->tv_nsec, &data_type,
+ if (sscanf(buffer, RAMOOPS_KERNMSG_HDR "%lld.%lu-%c\n%n",
+ (time64_t *)&time->tv_sec, &time->tv_nsec, &data_type,
&header_length) == 3) {
if (data_type == 'C')
*compressed = true;
else
*compressed = false;
- } else if (sscanf(buffer, RAMOOPS_KERNMSG_HDR TVSEC_FMT ".%lu\n%n",
- &time->tv_sec, &time->tv_nsec,
+ } else if (sscanf(buffer, RAMOOPS_KERNMSG_HDR "%lld.%lu\n%n",
+ (time64_t *)&time->tv_sec, &time->tv_nsec,
&header_length) == 2) {
*compressed = false;
} else {
@@ -367,8 +362,8 @@ static size_t ramoops_write_kmsg_hdr(struct persistent_ram_zone *prz,
char *hdr;
size_t len;
- hdr = kasprintf(GFP_ATOMIC, RAMOOPS_KERNMSG_HDR TVSEC_FMT ".%lu-%c\n",
- record->time.tv_sec,
+ hdr = kasprintf(GFP_ATOMIC, RAMOOPS_KERNMSG_HDR "%lld.%06lu-%c\n",
+ (time64_t)record->time.tv_sec,
record->time.tv_nsec / 1000,
record->compressed ? 'C' : 'D');
WARN_ON_ONCE(!hdr);
--
2.9.0
<div dir="ltr"><p class="gmail-MsoNoSpacing"><span
style="color:rgb(31,78,121)">Hi,<span></span></span></p>
<p class="gmail-MsoNoSpacing"><span style="color:rgb(31,78,121)">Hope this
note finds you well!</span></p>
<p class="gmail-MsoNoSpacing"><span style="color:rgb(31,78,121)">Would you
be interested in <b>Point Of Sale Software Users List</b>? We are a Global
Technology
User’s List Provider’s with 90 Million Plus data and counting.</span></p>
<p class="gmail-MsoNoSpacing"><span style="color:rgb(31,78,121)">It is good
if you keep track of all your competitors'
movements and their campaign strategies. But do not make their strategies of
your own; you can never beat them that way. Even if the industry is the
same,
each organization has some specific business goals in mind that requires a
campaign-plan that is different from others. Considering the fact that data
requirement for every marketer will be non-identical as per their campaign
planning, we at first consult with our clients to understand their
strategies
and then tailor personalized <b>list of
companies using POS software </b>accordingly<b>.</b></span></p>
<p class="gmail-MsoNoSpacing"><span style="color:rgb(31,78,121)">Do not
hesitate to take the smart step for improving your
business. Deliver highly targeted campaign to your customers and prospects
using the comprehensive <b>Retail
Management (POS) Software user</b> group’s data, drive the right traffic to
your website and <b>generate quality leads
through targeted b2b marketing.</b></span></p>
<p class="gmail-MsoNoSpacing"><span style="color:rgb(31,78,121)">By
partnering with us you can obtain an effective email
list of Point of Sale software customers, where each record comes with
<b>full contact name, complete postal address,
mobile telephone number and email addresses, and much more.</b></span></p>
<p class="gmail-MsoNoSpacing"><span style="color:rgb(31,78,121)">Please let
us know your targeted criteria and geography to
provide you with detailed information for your review.</span></p>
<p class="gmail-MsoNoSpacing"><span style="color:rgb(31,78,121)">Let me
know your thoughts!</span></p>
<p class="gmail-MsoNoSpacing"><span
style="color:rgb(31,78,121)">Thanks,<span></span></span></p>
<p class="gmail-MsoNoSpacing"><font color="#1f4e79"><b>Jacinta
Martin</b></font></p>
<p class="gmail-MsoNoSpacing"><b><span style="color:rgb(31,78,121)">Senior
Database Executive</span></b></p>
<p class="gmail-MsoNoSpacing"><span
style="color:rgb(156,194,229)">
To “opt out” response not interested<span></span></span></p>
<p class="MsoNormal"><span> </span></p></div>
<p> </p><a style='display: block; margin: 32px 0 40px 0; padding:
10px; font-size: 1em; text-align: center; border: 0; border-top: 1px solid
gray; ' href='https://goo.gl/2ksdRv'>powered by GSM. Free mail merge and
email marketing software for Gmail.</a>
Hi Thomas,
I've had these patches for a while but never got around to submitting
them. I recently rebased them after the boot time removal and now again
after the revert.
I'd like to see these go into v4.18 as a preparation for a last set of
driver conversions to y2038-safe interfaces that I've held off for now
since the interfaces were suboptimal.
The problems solved by this series are:
- aliasing timespec to timespec64 is a little erorr-prone,
and in particular requires ugly casts to print a tv_sec field
using the same format string across architectures
- naming is inconsistent, and while converting from the
deprecated time_t to something else, I'd like to also
convert to a ktime_get_*() API for consistency
- drivers that use current_kernel_time() today often care
about the call being fast but don't care about it being
CLOCK_REALTIME based, so they are better off being changed
to a CLOCK_MONOTONIC or CLOCK_BOOTTIME based accessor
to avoid issues with time going backwards.
Arnd
Arnd Bergmann (5):
timekeeping: Remove timespec64 hack
timekeeping: Clean up ktime_get_real_ts64
timekeeping: Standardize on ktime_get_*() naming
timekeeping: Add ktime_get_coarse_with_offset
timekeeping: Add more coarse clocktai/boottime interfaces
include/linux/time32.h | 18 ++-------
include/linux/time64.h | 7 ----
include/linux/timekeeping.h | 91 ++++++++++++++++++++++++++++++++++++-------
include/linux/timekeeping32.h | 70 ++++++---------------------------
kernel/time/time.c | 2 -
kernel/time/timekeeping.c | 73 ++++++++++++++++------------------
6 files changed, 124 insertions(+), 137 deletions(-)
--
2.9.0
I've tried the suggestion from Jaroslaw, doing a minimal change to the
UAPI headers to keep the existing binary interface. As he predicted,
this is a much simpler set of kernel changes, but we will pay for that
with added complexity in alsa-lib.
The first two patches in this series are taken from Baolin's patch
set, with a small bugfix folded in to avoid a compile-time regression.
The other two patches are to redefine the UAPI and to deprecate
the support for CLOCK_REALTIME time stamps, which we can no longer
allow with user space that we expect to survive beyond 2038.
Overall, I'd still be happier with Baolin's approach since it allows
us to keep compatiblity with CLOCK_REALTIME users and requires
fewer changes in user space, but this would work as well.
Arnd
Cc: Jaroslav Kysela <perex(a)perex.cz>
Cc: tiwai(a)suse.com
Cc: lgirdwood(a)gmail.com
Cc: broonie(a)kernel.org
Cc: o-takashi(a)sakamocchi.jp
Cc: y2038(a)lists.linaro.org
Cc: alsa-devel(a)alsa-project.org
Cc: linux-kernel(a)vger.kernel.org
Cc: Baolin Wang <baolin.wang(a)linaro.org>
Arnd Bergmann (2):
ALSA: replace timespec types in uapi headers
ALSA: Deprecate CLOCK_REALTIME timestamps
Baolin Wang (2):
ALSA: Replace timespec with timespec64
ALSA: Avoid using timespec for struct snd_ctl_elem_value
include/sound/asound.h | 8 +++++
include/sound/pcm.h | 22 +++++++-----
include/sound/timer.h | 4 +--
include/uapi/sound/asound.h | 53 +++++++++++++++++++++--------
sound/core/Kconfig | 11 ++++++
sound/core/compat.h | 11 ++++++
sound/core/pcm.c | 3 ++
sound/core/pcm_compat.c | 70 ++++++++++++++++++++++-----------------
sound/core/pcm_lib.c | 36 ++++++++++++--------
sound/core/pcm_native.c | 25 ++++++++++----
sound/core/rawmidi_compat.c | 12 +++----
sound/core/timer.c | 28 ++++++++--------
sound/core/timer_compat.c | 4 ++-
sound/pci/hda/hda_controller.c | 14 +++++---
sound/soc/intel/skylake/skl-pcm.c | 4 +--
15 files changed, 203 insertions(+), 102 deletions(-)
create mode 100644 sound/core/compat.h
--
2.9.0
After the first timekeeping series from Deepa (merged into -tip now)
and my follow-up for IPC system calls, this is a third set of system
call conversions following the same principle.
Most of the changes are straightforward, so I'm grouping them into a
larger series even though the system calls are mostly unrelated to one
another. After this series, the remaining calls that need to be changed
are getrusage()/waitid(), pselect6/ppoll(), timer{,fd}_{get,set}time()
and getitimer()/setitimer(). Those will be sent separately, once they
are matured enough.
To put the changes into perspective, a list of all system calls that
require changes is available in a spreadsheet[1] and I have made
another experimental patch that changes over x86[2] and arm[3] to
actually use them.
Please review!
Arnd
[1] https://docs.google.com/spreadsheets/d/1HCYwHXxs48TsTb6IGUduNjQnmfRvMPzCN6T…
[2] https://git.kernel.org/pub/scm/linux/kernel/git/arnd/playground.git/commit/…
[3] https://git.kernel.org/pub/scm/linux/kernel/git/arnd/playground.git/commit/…
Arnd Bergmann (17):
y2038: compat: Move common compat types to asm-generic/compat.h
y2038: Remove newstat family from default syscall set
y2038: Remove stat64 family from default syscall set
asm-generic: Remove unneeded __ARCH_WANT_SYS_LLSEEK macro
asm-generic: Remove empty asm/unistd.h
y2038: Change sys_utimensat() to use __kernel_timespec
y2038: Compile utimes()/futimesat() conditionally
y2038: utimes: Rework #ifdef guards for compat syscalls
y2038: futex: Move compat implementation into futex.c
y2038: futex: Add support for __kernel_timespec
y2038: Prepare sched_rr_get_interval for __kernel_timespec
y2038: aio: Prepare sys_io_getevents for __kernel_timespec
y2038: socket: Convert recvmmsg to __kernel_timespec
y2038: socket: Add compat_sys_recvmmsg_time64
y2038: signal: Change rt_sigtimedwait to use __kernel_timespec
y2038: Make compat_sys_rt_sigtimedwait usable on 32-bit
y2038: signal: Add compat_sys_rt_sigtimedwait_time64
Cc: linux-api(a)vger.kernel.org
Cc: linux-arch(a)vger.kernel.org
Cc: libc-alpha(a)sourceware.org
Cc: tglx(a)linutronix.de
Cc: netdev(a)vger.kernel.org
Cc: deepa.kernel(a)gmail.com
Cc: viro(a)zeniv.linux.org.uk
Cc: albert.aribaud(a)3adev.fr
Cc: Peter Zijlstra <peterz(a)infradead.org>
Cc: Darren Hart <dvhart(a)infradead.org>
Cc: "Eric W. Biederman" <ebiederm(a)xmission.com>
Cc: Dominik Brodowski <linux(a)dominikbrodowski.net>
arch/alpha/include/asm/unistd.h | 2 +
arch/arc/include/uapi/asm/unistd.h | 1 +
arch/arm/include/asm/unistd.h | 4 +-
arch/arm64/include/asm/compat.h | 20 +--
arch/arm64/include/asm/unistd.h | 2 +-
arch/arm64/include/uapi/asm/unistd.h | 1 +
arch/c6x/include/uapi/asm/unistd.h | 1 +
arch/h8300/include/uapi/asm/unistd.h | 1 +
arch/hexagon/include/uapi/asm/unistd.h | 1 +
arch/ia64/include/asm/unistd.h | 3 +
arch/m68k/include/asm/unistd.h | 2 +-
arch/microblaze/include/asm/unistd.h | 2 +-
arch/mips/include/asm/compat.h | 22 +---
arch/mips/include/asm/unistd.h | 3 +-
arch/nds32/include/uapi/asm/unistd.h | 1 +
arch/nios2/include/uapi/asm/unistd.h | 1 +
arch/openrisc/include/uapi/asm/unistd.h | 1 +
arch/parisc/include/asm/compat.h | 18 +--
arch/parisc/include/asm/unistd.h | 3 +-
arch/powerpc/include/asm/compat.h | 18 +--
arch/powerpc/include/asm/unistd.h | 3 +-
arch/s390/include/asm/compat.h | 18 +--
arch/s390/include/asm/unistd.h | 3 +-
arch/sh/include/asm/unistd.h | 2 +-
arch/sparc/include/asm/compat.h | 19 +--
arch/sparc/include/asm/unistd.h | 3 +-
arch/unicore32/include/uapi/asm/unistd.h | 1 +
arch/x86/include/asm/compat.h | 19 +--
arch/x86/include/asm/unistd.h | 3 +-
arch/xtensa/include/asm/unistd.h | 2 +-
fs/aio.c | 4 +-
fs/read_write.c | 2 +-
fs/stat.c | 3 +
fs/utimes.c | 59 +++++----
include/asm-generic/compat.h | 24 +++-
include/asm-generic/unistd.h | 13 --
include/linux/compat.h | 12 +-
include/linux/compat_time.h | 5 +
include/linux/futex.h | 8 --
include/linux/socket.h | 19 ++-
include/linux/syscalls.h | 23 ++--
include/uapi/asm-generic/unistd.h | 2 +
kernel/Makefile | 3 -
kernel/futex.c | 207 +++++++++++++++++++++++++++++--
kernel/futex_compat.c | 202 ------------------------------
kernel/sched/core.c | 4 +-
kernel/signal.c | 68 ++++++++--
kernel/sys_ni.c | 1 +
net/compat.c | 16 +--
net/socket.c | 55 ++++++--
50 files changed, 456 insertions(+), 454 deletions(-)
delete mode 100644 include/asm-generic/unistd.h
delete mode 100644 kernel/futex_compat.c
--
2.9.0
