| Message ID | 20210510150413.59356-1-andriy.shevchenko@linux.intel.com |
|---|---|
| State | New |
| Headers | show |
| Series | [v1,1/4] lib/vsprintf: Allow to override date and time separator | expand |
On Mon, May 10, 2021 at 05:05:46PM -0700, Doug Anderson wrote: > On Mon, May 10, 2021 at 8:04 AM Andy Shevchenko > <andriy.shevchenko@linux.intel.com> wrote: > > > > Use %ptTs instead of open-coded variant to print contents > > of time64_t type in human readable form. > > > > Cc: Jason Wessel <jason.wessel@windriver.com> > > Cc: Daniel Thompson <daniel.thompson@linaro.org> > > Cc: kgdb-bugreport@lists.sourceforge.net > > Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> > > --- > > kernel/debug/kdb/kdb_main.c | 9 +-------- > > 1 file changed, 1 insertion(+), 8 deletions(-) > > I kinda doubt anyone would really care if we just switched kdb to just > the old "%ptT". Probably no machines are parsing this string. Formally it's an ABI. > ...but in any case, now that the nifty new format is there we might as > well use it. Thus: > > Reviewed-by: Douglas Anderson <dianders@chromium.org> Thanks! -- With Best Regards, Andy Shevchenko
On Mon 2021-05-10 18:04:10, Andy Shevchenko wrote: > ISO 8601 defines 'T' as a separator between date and time. Though, > some ABIs use time and date with ' ' separator instead. > > Add a flavour to the %pt specifier to override default separator. > > Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> > --- > Documentation/core-api/printk-formats.rst | 6 +++++- > lib/test_printf.c | 5 +++++ > lib/vsprintf.c | 19 ++++++++++++++++--- > 3 files changed, 26 insertions(+), 4 deletions(-) > > diff --git a/Documentation/core-api/printk-formats.rst b/Documentation/core-api/printk-formats.rst > index f063a384c7c8..bc85fd4685e7 100644 > --- a/Documentation/core-api/printk-formats.rst > +++ b/Documentation/core-api/printk-formats.rst > @@ -514,9 +514,10 @@ Time and date > :: > > %pt[RT] YYYY-mm-ddTHH:MM:SS > + %pt[RT]s YYYY-mm-dd HH:MM:SS > %pt[RT]d YYYY-mm-dd > %pt[RT]t HH:MM:SS > - %pt[RT][dt][r] > + %pt[RT][dt][rs] Sigh, we do not have clear rules what [xy] means. It might be: + always use one of them + optionally use one of them + always use any of them + optionally use any of them %pt[RT][dt][rs] is a great mix: + R or T is required, the rest is optional + 'd' or 't' can be used but both together are not supported + any variant of 'r' and 's' is supported including various ordering Honestly, I do not know about any magic solution that might make it easier to understand these monster modifiers. Well, what about using the following at least in this case: %pt[RT][dt][r][s] It might help to understand that both 'r' and 's' can be used at the same time. An attempt to distinguishing all the possibilities might be: %pt{RT}[{dt}][r][s] where [] means that it is optional and {} means one of them must be chosen. But I am not sure if it really makes the life easier. Anyway, this would be for another patch that updates the entire printk-formats.rst. > > For printing date and time as represented by:: > > @@ -528,6 +529,9 @@ in human readable format. > By default year will be incremented by 1900 and month by 1. > Use %pt[RT]r (raw) to suppress this behaviour. > > +The %pt[RT]s (space) will override ISO 8601 by using ' ' instead of 'T' > +between date and time. It won't have any effect when date or time is omitted. > + > Passed by reference. > > struct clk > diff --git a/lib/vsprintf.c b/lib/vsprintf.c > index f0c35d9b65bf..5f36c7a43cdc 100644 > --- a/lib/vsprintf.c > +++ b/lib/vsprintf.c > @@ -1834,7 +1834,8 @@ char *rtc_str(char *buf, char *end, const struct rtc_time *tm, > struct printf_spec spec, const char *fmt) > { > bool have_t = true, have_d = true; > - bool raw = false; > + bool raw = false, space = false; > + bool found = true; > int count = 2; > > if (check_pointer(&buf, end, tm, spec)) > @@ -1851,14 +1852,26 @@ char *rtc_str(char *buf, char *end, const struct rtc_time *tm, > break; > } > > - raw = fmt[count] == 'r'; > + do { > + switch (fmt[count++]) { > + case 'r': > + raw = true; > + break; > + case 's': > + space = true; > + break; > + default: > + found = false; > + break; > + } > + } while (found); > > if (have_d) > buf = date_str(buf, end, tm, raw); > if (have_d && have_t) { > /* Respect ISO 8601 */ The comment is slightly misleding now. What about something like? /* 'T' by ISO 8601. */ Or maybe call the variable: iso_8601, remove the comment, and invert the logic: bool iso_8601 = true; case 's': iso_8601 = false; break; *buf = iso_8601 ? 'T' : ' '; > if (buf < end) > - *buf = 'T'; > + *buf = space ? ' ' : 'T'; > buf++; > } > if (have_t) Best Regards, Petr
On Mon 2021-05-10 18:04:11, Andy Shevchenko wrote: > Use %ptTs instead of open-coded variant to print contents > of time64_t type in human readable form. > > Cc: Jason Wessel <jason.wessel@windriver.com> > Cc: Daniel Thompson <daniel.thompson@linaro.org> > Cc: kgdb-bugreport@lists.sourceforge.net > Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Reviewed-by: Petr Mladek <pmladek@suse.com> Best Regards, Petr
On Mon 2021-05-10 18:04:13, Andy Shevchenko wrote: > Use %ptTs instead of open coded variant to print contents > of time64_t type in human readable form. > > Cc: Thierry Reding <thierry.reding@gmail.com> > Cc: Jonathan Hunter <jonathanh@nvidia.com> > Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Reviewed-by: Petr Mladek <pmladek@suse.com> Best Regards, Petr
On Tue, May 11, 2021 at 04:14:56PM +0200, Petr Mladek wrote: > On Mon 2021-05-10 18:04:10, Andy Shevchenko wrote: > > ISO 8601 defines 'T' as a separator between date and time. Though, > > some ABIs use time and date with ' ' separator instead. > > > > Add a flavour to the %pt specifier to override default separator. ... > > %pt[RT] YYYY-mm-ddTHH:MM:SS > > + %pt[RT]s YYYY-mm-dd HH:MM:SS > > %pt[RT]d YYYY-mm-dd > > %pt[RT]t HH:MM:SS > > - %pt[RT][dt][r] > > + %pt[RT][dt][rs] > > Sigh, we do not have clear rules what [xy] means. It might be: > > + always use one of them > + optionally use one of them > + always use any of them > + optionally use any of them > > %pt[RT][dt][rs] is a great mix: > > + R or T is required, the rest is optional > + 'd' or 't' can be used but both together are not supported > + any variant of 'r' and 's' is supported including various ordering > > Honestly, I do not know about any magic solution that might make it > easier to understand these monster modifiers. > > Well, what about using the following at least in this case: > > %pt[RT][dt][r][s] > > It might help to understand that both 'r' and 's' can be used at the > same time. This is the case, yes, thanks for catching it. > An attempt to distinguishing all the possibilities might be: > > %pt{RT}[{dt}][r][s] > > where [] means that it is optional and {} means one of them must be > chosen. But I am not sure if it really makes the life easier. Anyway, > this would be for another patch that updates the entire printk-formats.rst. No, this is not the case, the d and t can go in any combinations: none, d, t, dt, or td. > > > For printing date and time as represented by:: > > > > @@ -528,6 +529,9 @@ in human readable format. > > By default year will be incremented by 1900 and month by 1. > > Use %pt[RT]r (raw) to suppress this behaviour. > > > > +The %pt[RT]s (space) will override ISO 8601 by using ' ' instead of 'T' > > +between date and time. It won't have any effect when date or time is omitted. ... > > + do { > > + switch (fmt[count++]) { > > + case 'r': > > + raw = true; > > + break; > > + case 's': > > + space = true; > > + break; > > + default: > > + found = false; > > + break; > > + } > > + } while (found); > > > > if (have_d) > > buf = date_str(buf, end, tm, raw); > > if (have_d && have_t) { > > /* Respect ISO 8601 */ > > The comment is slightly misleding now. What about something like? > > /* 'T' by ISO 8601. */ > > Or maybe call the variable: iso_8601, remove the comment, and > invert the logic: Okay, I will think how to improve, thanks! > bool iso_8601 = true; > > case 's': > iso_8601 = false; > break; > > *buf = iso_8601 ? 'T' : ' '; > > > if (buf < end) > > - *buf = 'T'; > > + *buf = space ? ' ' : 'T'; > > buf++; > > } -- With Best Regards, Andy Shevchenko
diff --git a/Documentation/core-api/printk-formats.rst b/Documentation/core-api/printk-formats.rst index f063a384c7c8..bc85fd4685e7 100644 --- a/Documentation/core-api/printk-formats.rst +++ b/Documentation/core-api/printk-formats.rst @@ -514,9 +514,10 @@ Time and date :: %pt[RT] YYYY-mm-ddTHH:MM:SS + %pt[RT]s YYYY-mm-dd HH:MM:SS %pt[RT]d YYYY-mm-dd %pt[RT]t HH:MM:SS - %pt[RT][dt][r] + %pt[RT][dt][rs] For printing date and time as represented by:: @@ -528,6 +529,9 @@ in human readable format. By default year will be incremented by 1900 and month by 1. Use %pt[RT]r (raw) to suppress this behaviour. +The %pt[RT]s (space) will override ISO 8601 by using ' ' instead of 'T' +between date and time. It won't have any effect when date or time is omitted. + Passed by reference. struct clk diff --git a/lib/test_printf.c b/lib/test_printf.c index ec0d5976bb69..8ac71aee46af 100644 --- a/lib/test_printf.c +++ b/lib/test_printf.c @@ -528,6 +528,11 @@ time_and_date(void) test("0119-00-04T15:32:23", "%ptTr", &t); test("15:32:23|2019-01-04", "%ptTt|%ptTd", &t, &t); test("15:32:23|0119-00-04", "%ptTtr|%ptTdr", &t, &t); + + test("2019-01-04 15:32:23", "%ptTs", &t); + test("0119-00-04 15:32:23", "%ptTsr", &t); + test("15:32:23|2019-01-04", "%ptTts|%ptTds", &t, &t); + test("15:32:23|0119-00-04", "%ptTtrs|%ptTdrs", &t, &t); } static void __init diff --git a/lib/vsprintf.c b/lib/vsprintf.c index f0c35d9b65bf..5f36c7a43cdc 100644 --- a/lib/vsprintf.c +++ b/lib/vsprintf.c @@ -1834,7 +1834,8 @@ char *rtc_str(char *buf, char *end, const struct rtc_time *tm, struct printf_spec spec, const char *fmt) { bool have_t = true, have_d = true; - bool raw = false; + bool raw = false, space = false; + bool found = true; int count = 2; if (check_pointer(&buf, end, tm, spec)) @@ -1851,14 +1852,26 @@ char *rtc_str(char *buf, char *end, const struct rtc_time *tm, break; } - raw = fmt[count] == 'r'; + do { + switch (fmt[count++]) { + case 'r': + raw = true; + break; + case 's': + space = true; + break; + default: + found = false; + break; + } + } while (found); if (have_d) buf = date_str(buf, end, tm, raw); if (have_d && have_t) { /* Respect ISO 8601 */ if (buf < end) - *buf = 'T'; + *buf = space ? ' ' : 'T'; buf++; } if (have_t)
ISO 8601 defines 'T' as a separator between date and time. Though, some ABIs use time and date with ' ' separator instead. Add a flavour to the %pt specifier to override default separator. Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> --- Documentation/core-api/printk-formats.rst | 6 +++++- lib/test_printf.c | 5 +++++ lib/vsprintf.c | 19 ++++++++++++++++--- 3 files changed, 26 insertions(+), 4 deletions(-)