| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
| |
clang isn't smart enough to recognise the pattern:
if (foo) { ...} { else orig_year = ... }
...
if (!foo) { ... use orig_year .. }
So just unconditionally initialise orig_year.
|
|
|
|
|
|
| |
We only ever assign from `struct tm` to `struct TM`, not the other way
around, so making this const regardless of what `struct tm` has should
be safe.
|
| |
|
|
|
|
|
|
|
|
|
| |
The gmtime_r() and localtime_r() functions are not thread safe with
regards to changes in the environment and locale while executing. This
protects them with a mutex, but only from other functions that also
participate in using that mutex. Future commits will change everything
in the perl core to do that, but there is still a hole if some XS
writer, or alien library doesn't use the mutex.
|
|
|
|
|
|
|
|
|
|
| |
This macro is now only used once. It is accessible outside perl because
it is in time64.h, but isn't used in cpan. Remove it and place its
expansion in time64.c.
Part of the expansion called another function. That function can be
removed and the expansion simplified, since gmtime() is now
automatically converted into gmtime_r() if available and needed.
|
|
|
|
|
|
|
|
|
| |
This macro is now only used once, and was the only caller of a helper
function. Thus both can be removed and inlined, simplifying things.
Further, most of the function's functionality can be removed, since it
was an attempt to simulate localtime_r(), which is now automatically
furnished by reentrant.[ch] if needed.
|
|
|
|
|
| |
Indent code blocks created by the previous commit; a few more
white-space adjustments.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
There are two main cases in Perl_localtime64_r(), call them A and B.
Prior to this commit, it was structured
if (A) {
codeA
return;
}
codeB
return;
The problem this commit solves is that there is common stuff in codeA
and codeB that will now only be specified once. Currently that common
stuff is minimal, and so acceptable, but the next few commits will make
the common stuff more complex, so making it be only once is helpful.
|
|
|
|
| |
This is in preparation for it being used in more than one place
|
|
|
|
|
|
|
|
|
|
| |
* The struct TM * pointers to S_check_tm and S_timegm64 functions is
not modified, so make it be a const and let the compiler know it won't
get modified.
* Localize variables to their innermost scopes
* Consting various local variables
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
In order to support the C<vmsish 'time'> pragma, the wrappers for
gmtime and localtime need thread context, which in the case of
localtime was removed in 23491f1decae17401c5. Add it back but
make it VMS-specific and turn the comment indicating why it's
there into a more direct statement.
|
| |
|
|
|
|
|
| |
The data in the const static arrays can be expressed with smaller
datatypes. This reduces binary size slightly.
|
|
|
|
|
|
| |
The original version just zeroed dest if src == NULL, but that code path
was never used. (gcc -Os inlined the function and optimized the test
away anyway.)
|
|
|
|
|
|
|
|
|
| |
All callers of S_copy_little_tm_to_big_TM pass a true variable src.
Checking for null is pointless. On Win32 32 bit VC 2003 -O1 -GL,
perl517.dll's .text section went from 0xCO13F to 0xC012F bytes long. It can
be argued that the compiler should have figured this out on its own, but VC
for whatever reason didn't. Also pretty the indenting and align the
assignments. The null check blames to commit 806a119aef .
|
|
|
|
|
| |
Remove either unused dTHXes, or remove dTHXes where a nocontext func can
be used instead. Smaller/faster machine code is the result.
|
|
|
|
|
|
|
|
|
| |
# New Ticket Created by (Peter J. Acklam)
# Please include the string: [perl #81904]
# in the subject line of all future correspondence about this issue.
# <URL: http://rt.perl.org/rt3/Ticket/Display.html?id=81904 >
Signed-off-by: Abigail <abigail@abigail.be>
|
|
|
|
| |
to a char *.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Michael G Schwern wrote:
> Anyhow, I'm willing to let that drop, but I want figure out A) why its
> failing so early because it might be masking a bug, this is not
> predicted behavior, and most importantly B) to get it to error, or at
> least warn.
>
> I know B is possible, I did it for Time::y2038. I'll figure that part
> out and try to dig into A. I would like to call B a blocker for 5.12.
Found the bug, patch attached.
v_tm_tday is an NV, so casting time to an integer will cause an overflow
2**31 days after 1970 or somewhere between 2**47 and 2**48. Not sure why the
compiler didn't warn about that.
This makes it work until the year overflows.
As for putting in a warning, there's two options:
1) A few days before the 32 bit year overflows.
2) At some reasonable point before there's too much floating point inaccuracy.
The possibility of long doubles will make #2 difficult to nail down, but I'll
be happy with +/- 2**53.
--
24. Must not tell any officer that I am smarter than they are, especially
if it's true.
-- The 213 Things Skippy Is No Longer Allowed To Do In The U.S. Army
http://skippyslist.com/list/
From 9e940e610ac5b2fbb09a554f505963094a4a0745 Mon Sep 17 00:00:00 2001
From: Michael G. Schwern <schwern@pobox.com>
Date: Sun, 31 Jan 2010 02:24:50 -0800
Subject: [PATCH] Type conversion bug in gmtime64 that was causing it to crap out around 2**48
Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
|
| |
|
|
|
|
|
|
|
| |
This means it should run on anything that does not have a 64-bit
integer type available but does have a double. Presumably this
includes platforms that define PERL_MICRO, so we now use the
same extended time implementation for everything that runs Perl.
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
All of the time64 functions are only visible within pp_sys.c and are not
part of the public API, so make them all static and give them the S_ prefix.
A side effect of this is that we now unambiguously use S_localtime64_r and
S_gmtime64_r and should no longer collide with system-supplied versions, if
any.
One possible future direction is that if and when Configure detects any
system-supplied *time64_r functions, their prototypes, and their valid
time ranges, we could optionally select those instead of the home-grown
ones.
Another possible future direction is that we could s/S_/Perl_/ and make
the *time64_r functions part of the public API in some post-5.10.x
release. Currently they are only exposed to the outside world via
Perl_pp_gmtime.
Also gave some TRACE macros more specific names.
|
|
|
|
|
|
|
| |
Silence the following build warning:
time64.c:169: warning: 'check_tm' defined but not used
Message-ID: <1ff86f510901081219q4f5f25a3od89c4e5ac03e8664@mail.gmail.com>
|
| |
|
|
|
|
| |
They definitely need it on VMS.
|
| |
|
|
|
|
|
|
| |
Fix trailing #endif.
Remove C99 macro.
|
|
|
|
|
|
|
|
|
|
|
| |
Add trace code.
Fix implied negative time in localtie64_r(). This fixes Windows.
Put in some more tests around small negative and positive times to
try and catch the above.
Explain the loss of accuracy due to use of doubles in perlport.
|
|
|
|
|
|
|
|
|
|
|
|
| |
Configuration information split out into its own header.
Added files to MANIFEST
Turn off USE_SYSTEM_GMTIME, ours is more reliable and possibly faster.
Fix type warnings found on Windows.
Remove unnecessary use of floor() and ceil().
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
LOCALTIME_MAX to 0x7fff573e.
Update from y2038.
Use the new TM64 struct so years can go out past y2**31
Defines a Year type to avoid converting years to ints.
Remove the TIMGM work around code, using timegm64() is fine and it
saves us from having to convert from TM to tm.
Make functions private with static rather than the _foo convention.
Even faster for distant dates.
|
|
file isn't about just localtime() anymore.
|