This adds the implementation of misc operations for exfat.
Signed-off-by: Namjae Jeon <[email protected]>
Signed-off-by: Sungjong Seo <[email protected]>
---
fs/exfat/misc.c | 253 ++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 253 insertions(+)
create mode 100644 fs/exfat/misc.c
diff --git a/fs/exfat/misc.c b/fs/exfat/misc.c
new file mode 100644
index 000000000000..7f533bcb3b3f
--- /dev/null
+++ b/fs/exfat/misc.c
@@ -0,0 +1,253 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Written 1992,1993 by Werner Almesberger
+ * 22/11/2000 - Fixed fat_date_unix2dos for dates earlier than 01/01/1980
+ * and date_dos2unix for date==0 by Igor Zhbanov([email protected])
+ * Copyright (C) 2012-2013 Samsung Electronics Co., Ltd.
+ */
+
+#include <linux/time.h>
+#include <linux/fs.h>
+#include <linux/slab.h>
+#include <linux/buffer_head.h>
+
+#include "exfat_raw.h"
+#include "exfat_fs.h"
+
+/*
+ * exfat_fs_error reports a file system problem that might indicate fa data
+ * corruption/inconsistency. Depending on 'errors' mount option the
+ * panic() is called, or error message is printed FAT and nothing is done,
+ * or filesystem is remounted read-only (default behavior).
+ * In case the file system is remounted read-only, it can be made writable
+ * again by remounting it.
+ */
+void __exfat_fs_error(struct super_block *sb, int report, const char *fmt, ...)
+{
+ struct exfat_mount_options *opts = &EXFAT_SB(sb)->options;
+ va_list args;
+ struct va_format vaf;
+
+ if (report) {
+ va_start(args, fmt);
+ vaf.fmt = fmt;
+ vaf.va = &args;
+ exfat_msg(sb, KERN_ERR, "error, %pV\n", &vaf);
+ va_end(args);
+ }
+
+ if (opts->errors == EXFAT_ERRORS_PANIC) {
+ panic("exFAT-fs (%s): fs panic from previous error\n",
+ sb->s_id);
+ } else if (opts->errors == EXFAT_ERRORS_RO && !sb_rdonly(sb)) {
+ sb->s_flags |= SB_RDONLY;
+ exfat_msg(sb, KERN_ERR, "Filesystem has been set read-only");
+ }
+}
+
+/*
+ * exfat_msg() - print preformated EXFAT specific messages.
+ * All logs except what uses exfat_fs_error() should be written by exfat_msg()
+ */
+void exfat_msg(struct super_block *sb, const char *level, const char *fmt, ...)
+{
+ struct va_format vaf;
+ va_list args;
+
+ va_start(args, fmt);
+ vaf.fmt = fmt;
+ vaf.va = &args;
+ /* level means KERN_ pacility level */
+ printk("%sexFAT-fs (%s): %pV\n", level, sb->s_id, &vaf);
+ va_end(args);
+}
+
+/* <linux/time.h> externs sys_tz
+ * extern struct timezone sys_tz;
+ */
+#define UNIX_SECS_1980 315532800L
+
+#if BITS_PER_LONG == 64
+#define UNIX_SECS_2108 4354819200L
+#endif
+
+/* days between 1970/01/01 and 1980/01/01 (2 leap days) */
+#define DAYS_DELTA_DECADE (365 * 10 + 2)
+/* 120 (2100 - 1980) isn't leap year */
+#define NO_LEAP_YEAR_2100 (120)
+#define IS_LEAP_YEAR(y) (!((y) & 0x3) && (y) != NO_LEAP_YEAR_2100)
+
+#define SECS_PER_MIN (60)
+#define SECS_PER_HOUR (60 * SECS_PER_MIN)
+#define SECS_PER_DAY (24 * SECS_PER_HOUR)
+
+#define MAKE_LEAP_YEAR(leap_year, year) \
+ do { \
+ /* 2100 isn't leap year */ \
+ if (unlikely(year > NO_LEAP_YEAR_2100)) \
+ leap_year = ((year + 3) / 4) - 1; \
+ else \
+ leap_year = ((year + 3) / 4); \
+ } while (0)
+
+/* Linear day numbers of the respective 1sts in non-leap years. */
+static time_t accum_days_in_year[] = {
+ /* Month : N 01 02 03 04 05 06 07 08 09 10 11 12 */
+ 0, 0, 31, 59, 90, 120, 151, 181, 212, 243, 273, 304, 334, 0, 0, 0,
+};
+
+#define TIMEZONE_SEC(x) ((x) * 15 * SECS_PER_MIN)
+/* Convert a FAT time/date pair to a UNIX date (seconds since 1 1 70). */
+void exfat_time_fat2unix(struct exfat_sb_info *sbi, struct timespec64 *ts,
+ struct exfat_date_time *tp)
+{
+ time_t year = tp->year;
+ time_t ld; /* leap day */
+
+ MAKE_LEAP_YEAR(ld, year);
+
+ if (IS_LEAP_YEAR(year) && (tp->month) > 2)
+ ld++;
+
+ ts->tv_sec = tp->second + tp->minute * SECS_PER_MIN +
+ tp->hour * SECS_PER_HOUR +
+ (year * 365 + ld + accum_days_in_year[tp->month] +
+ (tp->day - 1) + DAYS_DELTA_DECADE) * SECS_PER_DAY;
+
+ ts->tv_nsec = 0;
+
+ /* Treat as local time */
+ if (!tp->timezone.valid) {
+ ts->tv_sec += sys_tz.tz_minuteswest * SECS_PER_MIN;
+ return;
+ }
+
+ /* Treat as UTC time, but need to adjust timezone to UTC0 */
+ if (tp->timezone.off <= 0x3F)
+ ts->tv_sec -= TIMEZONE_SEC(tp->timezone.off);
+ else /* 0x40 <= (tp->timezone & 0x7F) <=0x7F */
+ ts->tv_sec += TIMEZONE_SEC(0x80 - tp->timezone.off);
+}
+
+#define TIMEZONE_CUR_OFFSET() ((sys_tz.tz_minuteswest / (-15)) & 0x7F)
+/* Convert linear UNIX date to a FAT time/date pair. */
+void exfat_time_unix2fat(struct exfat_sb_info *sbi, struct timespec64 *ts,
+ struct exfat_date_time *tp)
+{
+ time_t second = ts->tv_sec;
+ time_t day, month, year;
+ time_t ld; /* leap day */
+
+ /* Treats as local time with proper time */
+ second -= sys_tz.tz_minuteswest * SECS_PER_MIN;
+ tp->timezone.valid = 1;
+ tp->timezone.off = TIMEZONE_CUR_OFFSET();
+
+ /* Jan 1 GMT 00:00:00 1980. But what about another time zone? */
+ if (second < UNIX_SECS_1980) {
+ tp->second = 0;
+ tp->minute = 0;
+ tp->hour = 0;
+ tp->day = 1;
+ tp->month = 1;
+ tp->year = 0;
+ return;
+ }
+
+ if (second >= UNIX_SECS_2108) {
+ tp->second = 59;
+ tp->minute = 59;
+ tp->hour = 23;
+ tp->day = 31;
+ tp->month = 12;
+ tp->year = 127;
+ return;
+ }
+
+ day = second / SECS_PER_DAY - DAYS_DELTA_DECADE;
+ year = day / 365;
+
+ MAKE_LEAP_YEAR(ld, year);
+ if (year * 365 + ld > day)
+ year--;
+
+ MAKE_LEAP_YEAR(ld, year);
+ day -= year * 365 + ld;
+
+ if (IS_LEAP_YEAR(year) && day == accum_days_in_year[3]) {
+ month = 2;
+ } else {
+ if (IS_LEAP_YEAR(year) && day > accum_days_in_year[3])
+ day--;
+ for (month = 1; month < 12; month++) {
+ if (accum_days_in_year[month + 1] > day)
+ break;
+ }
+ }
+ day -= accum_days_in_year[month];
+
+ tp->second = second % SECS_PER_MIN;
+ tp->minute = (second / SECS_PER_MIN) % 60;
+ tp->hour = (second / SECS_PER_HOUR) % 24;
+ tp->day = day + 1;
+ tp->month = month;
+ tp->year = year;
+}
+
+struct exfat_timestamp *exfat_tm_now(struct exfat_sb_info *sbi,
+ struct exfat_timestamp *tp)
+{
+ struct timespec64 ts;
+ struct exfat_date_time dt;
+
+ ktime_get_real_ts64(&ts);
+ exfat_time_unix2fat(sbi, &ts, &dt);
+
+ tp->year = dt.year;
+ tp->mon = dt.month;
+ tp->day = dt.day;
+ tp->hour = dt.hour;
+ tp->min = dt.minute;
+ tp->sec = dt.second;
+ tp->tz.value = dt.timezone.value;
+
+ return tp;
+}
+
+unsigned short exfat_calc_chksum_2byte(void *data, int len,
+ unsigned short chksum, int type)
+{
+ int i;
+ unsigned char *c = (unsigned char *)data;
+
+ for (i = 0; i < len; i++, c++) {
+ if (((i == 2) || (i == 3)) && (type == CS_DIR_ENTRY))
+ continue;
+ chksum = (((chksum & 1) << 15) | ((chksum & 0xFFFE) >> 1)) +
+ (unsigned short)*c;
+ }
+ return chksum;
+}
+
+void exfat_update_bh(struct super_block *sb, struct buffer_head *bh, int sync)
+{
+ set_bit(EXFAT_SB_DIRTY, &EXFAT_SB(sb)->s_state);
+ set_buffer_uptodate(bh);
+ mark_buffer_dirty(bh);
+
+ if (sync)
+ sync_dirty_buffer(bh);
+}
+
+void exfat_chain_set(struct exfat_chain *ec, unsigned int dir,
+ unsigned int size, unsigned char flags)
+{
+ ec->dir = dir;
+ ec->size = size;
+ ec->flags = flags;
+}
+
+void exfat_chain_dup(struct exfat_chain *dup, struct exfat_chain *ec)
+{
+ return exfat_chain_set(dup, ec->dir, ec->size, ec->flags);
+}
--
2.17.1
On Thursday 02 January 2020 16:20:32 Namjae Jeon wrote:
> This adds the implementation of misc operations for exfat.
>
> Signed-off-by: Namjae Jeon <[email protected]>
> Signed-off-by: Sungjong Seo <[email protected]>
> ---
> fs/exfat/misc.c | 253 ++++++++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 253 insertions(+)
> create mode 100644 fs/exfat/misc.c
>
> diff --git a/fs/exfat/misc.c b/fs/exfat/misc.c
> new file mode 100644
> index 000000000000..7f533bcb3b3f
> --- /dev/null
> +++ b/fs/exfat/misc.c
...
> +/* <linux/time.h> externs sys_tz
> + * extern struct timezone sys_tz;
> + */
> +#define UNIX_SECS_1980 315532800L
> +
> +#if BITS_PER_LONG == 64
> +#define UNIX_SECS_2108 4354819200L
> +#endif
...
> +#define TIMEZONE_CUR_OFFSET() ((sys_tz.tz_minuteswest / (-15)) & 0x7F)
> +/* Convert linear UNIX date to a FAT time/date pair. */
> +void exfat_time_unix2fat(struct exfat_sb_info *sbi, struct timespec64 *ts,
> + struct exfat_date_time *tp)
> +{
> + time_t second = ts->tv_sec;
> + time_t day, month, year;
> + time_t ld; /* leap day */
Question for other maintainers: Has kernel code already time_t defined
as 64bit? Or it is still just 32bit and 32bit systems and some time64_t
needs to be used? I remember that there was discussion about these
problems, but do not know if it was changed/fixed or not... Just a
pointer for possible Y2038 problem. As "ts" is of type timespec64, but
"second" of type time_t.
> +
> + /* Treats as local time with proper time */
> + second -= sys_tz.tz_minuteswest * SECS_PER_MIN;
> + tp->timezone.valid = 1;
> + tp->timezone.off = TIMEZONE_CUR_OFFSET();
> +
> + /* Jan 1 GMT 00:00:00 1980. But what about another time zone? */
> + if (second < UNIX_SECS_1980) {
> + tp->second = 0;
> + tp->minute = 0;
> + tp->hour = 0;
> + tp->day = 1;
> + tp->month = 1;
> + tp->year = 0;
> + return;
> + }
> +
> + if (second >= UNIX_SECS_2108) {
Hello, this code cause compile errors on 32bit systems as UNIX_SECS_2108
macro is not defined when BITS_PER_LONG == 32.
Value 4354819200 really cannot fit into 32bit signed integer, so you
should use 64bit signed integer. I would suggest to define this macro
value via LL not just L suffix (and it would work on both 32 and 64bit)
#define UNIX_SECS_2108 4354819200LL
> + tp->second = 59;
> + tp->minute = 59;
> + tp->hour = 23;
> + tp->day = 31;
> + tp->month = 12;
> + tp->year = 127;
> + return;
> + }
--
Pali Rohár
[email protected]
2020-01-02 18:19 GMT+09:00, Pali Rohár <[email protected]>:
> On Thursday 02 January 2020 16:20:32 Namjae Jeon wrote:
>> This adds the implementation of misc operations for exfat.
>>
>> Signed-off-by: Namjae Jeon <[email protected]>
>> Signed-off-by: Sungjong Seo <[email protected]>
>> ---
>> fs/exfat/misc.c | 253 ++++++++++++++++++++++++++++++++++++++++++++++++
>> 1 file changed, 253 insertions(+)
>> create mode 100644 fs/exfat/misc.c
>>
>> diff --git a/fs/exfat/misc.c b/fs/exfat/misc.c
>> new file mode 100644
>> index 000000000000..7f533bcb3b3f
>> --- /dev/null
>> +++ b/fs/exfat/misc.c
>
> ...
>
>> +/* <linux/time.h> externs sys_tz
>> + * extern struct timezone sys_tz;
>> + */
>> +#define UNIX_SECS_1980 315532800L
>> +
>> +#if BITS_PER_LONG == 64
>> +#define UNIX_SECS_2108 4354819200L
>> +#endif
>
> ...
>
>> +#define TIMEZONE_CUR_OFFSET() ((sys_tz.tz_minuteswest / (-15)) & 0x7F)
>> +/* Convert linear UNIX date to a FAT time/date pair. */
>> +void exfat_time_unix2fat(struct exfat_sb_info *sbi, struct timespec64
>> *ts,
>> + struct exfat_date_time *tp)
>> +{
>> + time_t second = ts->tv_sec;
>> + time_t day, month, year;
>> + time_t ld; /* leap day */
>
> Question for other maintainers: Has kernel code already time_t defined
> as 64bit? Or it is still just 32bit and 32bit systems and some time64_t
> needs to be used? I remember that there was discussion about these
> problems, but do not know if it was changed/fixed or not... Just a
> pointer for possible Y2038 problem. As "ts" is of type timespec64, but
> "second" of type time_t.
My bad, I will change it with time64_t.
>
>> +
>> + /* Treats as local time with proper time */
>> + second -= sys_tz.tz_minuteswest * SECS_PER_MIN;
>> + tp->timezone.valid = 1;
>> + tp->timezone.off = TIMEZONE_CUR_OFFSET();
>> +
>> + /* Jan 1 GMT 00:00:00 1980. But what about another time zone? */
>> + if (second < UNIX_SECS_1980) {
>> + tp->second = 0;
>> + tp->minute = 0;
>> + tp->hour = 0;
>> + tp->day = 1;
>> + tp->month = 1;
>> + tp->year = 0;
>> + return;
>> + }
>> +
>> + if (second >= UNIX_SECS_2108) {
>
> Hello, this code cause compile errors on 32bit systems as UNIX_SECS_2108
> macro is not defined when BITS_PER_LONG == 32.
>
> Value 4354819200 really cannot fit into 32bit signed integer, so you
> should use 64bit signed integer. I would suggest to define this macro
> value via LL not just L suffix (and it would work on both 32 and 64bit)
Okay.
>
> #define UNIX_SECS_2108 4354819200LL
>
>> + tp->second = 59;
>> + tp->minute = 59;
>> + tp->hour = 23;
>> + tp->day = 31;
>> + tp->month = 12;
>> + tp->year = 127;
>> + return;
>> + }
Okay, I will check it.
Thanks for your review!
>
> --
> Pali Rohár
> [email protected]
>
On Thursday 02 January 2020 20:30:03 Namjae Jeon wrote:
> 2020-01-02 18:19 GMT+09:00, Pali Rohár <[email protected]>:
> > On Thursday 02 January 2020 16:20:32 Namjae Jeon wrote:
> >> This adds the implementation of misc operations for exfat.
> >>
> >> Signed-off-by: Namjae Jeon <[email protected]>
> >> Signed-off-by: Sungjong Seo <[email protected]>
> >> ---
> >> fs/exfat/misc.c | 253 ++++++++++++++++++++++++++++++++++++++++++++++++
> >> 1 file changed, 253 insertions(+)
> >> create mode 100644 fs/exfat/misc.c
> >>
> >> diff --git a/fs/exfat/misc.c b/fs/exfat/misc.c
> >> new file mode 100644
> >> index 000000000000..7f533bcb3b3f
> >> --- /dev/null
> >> +++ b/fs/exfat/misc.c
> >
> > ...
> >
> >> +/* <linux/time.h> externs sys_tz
> >> + * extern struct timezone sys_tz;
> >> + */
> >> +#define UNIX_SECS_1980 315532800L
> >> +
> >> +#if BITS_PER_LONG == 64
> >> +#define UNIX_SECS_2108 4354819200L
> >> +#endif
> >
> > ...
> >
> >> +#define TIMEZONE_CUR_OFFSET() ((sys_tz.tz_minuteswest / (-15)) & 0x7F)
> >> +/* Convert linear UNIX date to a FAT time/date pair. */
> >> +void exfat_time_unix2fat(struct exfat_sb_info *sbi, struct timespec64
> >> *ts,
> >> + struct exfat_date_time *tp)
> >> +{
> >> + time_t second = ts->tv_sec;
> >> + time_t day, month, year;
> >> + time_t ld; /* leap day */
> >
> > Question for other maintainers: Has kernel code already time_t defined
> > as 64bit? Or it is still just 32bit and 32bit systems and some time64_t
> > needs to be used? I remember that there was discussion about these
> > problems, but do not know if it was changed/fixed or not... Just a
> > pointer for possible Y2038 problem. As "ts" is of type timespec64, but
> > "second" of type time_t.
> My bad, I will change it with time64_t.
Now I have looked into sources and time_t is just typedef from
__kernel_old_time_t type. So it looks like that time64_t is the type
which should be used in new code.
But somebody else should confirm this information.
> >
> >> +
> >> + /* Treats as local time with proper time */
> >> + second -= sys_tz.tz_minuteswest * SECS_PER_MIN;
> >> + tp->timezone.valid = 1;
> >> + tp->timezone.off = TIMEZONE_CUR_OFFSET();
> >> +
> >> + /* Jan 1 GMT 00:00:00 1980. But what about another time zone? */
> >> + if (second < UNIX_SECS_1980) {
> >> + tp->second = 0;
> >> + tp->minute = 0;
> >> + tp->hour = 0;
> >> + tp->day = 1;
> >> + tp->month = 1;
> >> + tp->year = 0;
> >> + return;
> >> + }
> >> +
> >> + if (second >= UNIX_SECS_2108) {
> >
> > Hello, this code cause compile errors on 32bit systems as UNIX_SECS_2108
> > macro is not defined when BITS_PER_LONG == 32.
> >
> > Value 4354819200 really cannot fit into 32bit signed integer, so you
> > should use 64bit signed integer. I would suggest to define this macro
> > value via LL not just L suffix (and it would work on both 32 and 64bit)
> Okay.
> >
> > #define UNIX_SECS_2108 4354819200LL
> >
> >> + tp->second = 59;
> >> + tp->minute = 59;
> >> + tp->hour = 23;
> >> + tp->day = 31;
> >> + tp->month = 12;
> >> + tp->year = 127;
> >> + return;
> >> + }
> Okay, I will check it.
> Thanks for your review!
> >
> > --
> > Pali Rohár
> > [email protected]
> >
--
Pali Rohár
[email protected]
On Thursday 02 January 2020 12:40:34 Pali Rohár wrote:
> On Thursday 02 January 2020 20:30:03 Namjae Jeon wrote:
> > 2020-01-02 18:19 GMT+09:00, Pali Rohár <[email protected]>:
> > > On Thursday 02 January 2020 16:20:32 Namjae Jeon wrote:
> > >> This adds the implementation of misc operations for exfat.
> > >>
> > >> Signed-off-by: Namjae Jeon <[email protected]>
> > >> Signed-off-by: Sungjong Seo <[email protected]>
> > >> ---
> > >> fs/exfat/misc.c | 253 ++++++++++++++++++++++++++++++++++++++++++++++++
> > >> 1 file changed, 253 insertions(+)
> > >> create mode 100644 fs/exfat/misc.c
> > >>
> > >> diff --git a/fs/exfat/misc.c b/fs/exfat/misc.c
> > >> new file mode 100644
> > >> index 000000000000..7f533bcb3b3f
> > >> --- /dev/null
> > >> +++ b/fs/exfat/misc.c
> > >
> > > ...
> > >
> > >> +/* <linux/time.h> externs sys_tz
> > >> + * extern struct timezone sys_tz;
> > >> + */
> > >> +#define UNIX_SECS_1980 315532800L
> > >> +
> > >> +#if BITS_PER_LONG == 64
> > >> +#define UNIX_SECS_2108 4354819200L
> > >> +#endif
> > >
> > > ...
> > >
> > >> +#define TIMEZONE_CUR_OFFSET() ((sys_tz.tz_minuteswest / (-15)) & 0x7F)
> > >> +/* Convert linear UNIX date to a FAT time/date pair. */
> > >> +void exfat_time_unix2fat(struct exfat_sb_info *sbi, struct timespec64
> > >> *ts,
> > >> + struct exfat_date_time *tp)
> > >> +{
> > >> + time_t second = ts->tv_sec;
> > >> + time_t day, month, year;
> > >> + time_t ld; /* leap day */
> > >
> > > Question for other maintainers: Has kernel code already time_t defined
> > > as 64bit? Or it is still just 32bit and 32bit systems and some time64_t
> > > needs to be used? I remember that there was discussion about these
> > > problems, but do not know if it was changed/fixed or not... Just a
> > > pointer for possible Y2038 problem. As "ts" is of type timespec64, but
> > > "second" of type time_t.
> > My bad, I will change it with time64_t.
>
> Now I have looked into sources and time_t is just typedef from
> __kernel_old_time_t type. So it looks like that time64_t is the type
> which should be used in new code.
>
> But somebody else should confirm this information.
>
> > >
> > >> +
> > >> + /* Treats as local time with proper time */
> > >> + second -= sys_tz.tz_minuteswest * SECS_PER_MIN;
> > >> + tp->timezone.valid = 1;
> > >> + tp->timezone.off = TIMEZONE_CUR_OFFSET();
> > >> +
> > >> + /* Jan 1 GMT 00:00:00 1980. But what about another time zone? */
> > >> + if (second < UNIX_SECS_1980) {
> > >> + tp->second = 0;
> > >> + tp->minute = 0;
> > >> + tp->hour = 0;
> > >> + tp->day = 1;
> > >> + tp->month = 1;
> > >> + tp->year = 0;
> > >> + return;
> > >> + }
> > >> +
> > >> + if (second >= UNIX_SECS_2108) {
> > >
> > > Hello, this code cause compile errors on 32bit systems as UNIX_SECS_2108
> > > macro is not defined when BITS_PER_LONG == 32.
> > >
> > > Value 4354819200 really cannot fit into 32bit signed integer, so you
> > > should use 64bit signed integer. I would suggest to define this macro
> > > value via LL not just L suffix (and it would work on both 32 and 64bit)
> > Okay.
> > >
> > > #define UNIX_SECS_2108 4354819200LL
> > >
> > >> + tp->second = 59;
> > >> + tp->minute = 59;
> > >> + tp->hour = 23;
> > >> + tp->day = 31;
> > >> + tp->month = 12;
> > >> + tp->year = 127;
> > >> + return;
> > >> + }
> > Okay, I will check it.
> > Thanks for your review!
Also I think that you should apply this Arnd's patch:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=44f6b40c225eb8e82eba8fd96d8f3fb843bc5a09
And maybe it is a good idea to look at applied staging patches:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/log/drivers/staging/exfat
--
Pali Rohár
[email protected]
> Also I think that you should apply this Arnd's patch:
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=44f6b40c225eb8e82eba8fd96d8f3fb843bc5a09
Okay:)
>
> And maybe it is a good idea to look at applied staging patches:
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/log/drivers/staging/exfat
I will check them.
Thanks a lot!
>
> --
> Pali Rohár
> [email protected]
>
On Wed, Jan 8, 2020 at 7:03 PM Christoph Hellwig <[email protected]> wrote:
>
> Arnd, can you review the exfat time handling, especially vs y2038
> related issues?
Sure, thanks for adding me to the loop
> On Thu, Jan 02, 2020 at 10:19:02AM +0100, Pali Rohár wrote:
> > On Thursday 02 January 2020 16:20:32 Namjae Jeon wrote:
> > > +#define TIMEZONE_CUR_OFFSET() ((sys_tz.tz_minuteswest / (-15)) & 0x7F)
> > > +/* Convert linear UNIX date to a FAT time/date pair. */
> > > +void exfat_time_unix2fat(struct exfat_sb_info *sbi, struct timespec64 *ts,
> > > + struct exfat_date_time *tp)
> > > +{
> > > + time_t second = ts->tv_sec;
> > > + time_t day, month, year;
> > > + time_t ld; /* leap day */
> >
> > Question for other maintainers: Has kernel code already time_t defined
> > as 64bit? Or it is still just 32bit and 32bit systems and some time64_t
> > needs to be used? I remember that there was discussion about these
> > problems, but do not know if it was changed/fixed or not... Just a
> > pointer for possible Y2038 problem. As "ts" is of type timespec64, but
> > "second" of type time_t.
I am actually very close to sending the patches to remove the time_t
definition from the kernel, at least in yesterday's version there were no
users.
exfat_time_unix2fat() seems to be a copy of the old fat_time_unix2fat()
that we fixed a while ago, please have a look at that implementation
based on time64_to_tm(), which avoids time_t.
Arnd
Arnd, can you review the exfat time handling, especially vs y2038
related issues?
On Thu, Jan 02, 2020 at 10:19:02AM +0100, Pali Roh?r wrote:
> On Thursday 02 January 2020 16:20:32 Namjae Jeon wrote:
> > This adds the implementation of misc operations for exfat.
> >
> > Signed-off-by: Namjae Jeon <[email protected]>
> > Signed-off-by: Sungjong Seo <[email protected]>
> > ---
> > fs/exfat/misc.c | 253 ++++++++++++++++++++++++++++++++++++++++++++++++
> > 1 file changed, 253 insertions(+)
> > create mode 100644 fs/exfat/misc.c
> >
> > diff --git a/fs/exfat/misc.c b/fs/exfat/misc.c
> > new file mode 100644
> > index 000000000000..7f533bcb3b3f
> > --- /dev/null
> > +++ b/fs/exfat/misc.c
>
> ...
>
> > +/* <linux/time.h> externs sys_tz
> > + * extern struct timezone sys_tz;
> > + */
> > +#define UNIX_SECS_1980 315532800L
> > +
> > +#if BITS_PER_LONG == 64
> > +#define UNIX_SECS_2108 4354819200L
> > +#endif
>
> ...
>
> > +#define TIMEZONE_CUR_OFFSET() ((sys_tz.tz_minuteswest / (-15)) & 0x7F)
> > +/* Convert linear UNIX date to a FAT time/date pair. */
> > +void exfat_time_unix2fat(struct exfat_sb_info *sbi, struct timespec64 *ts,
> > + struct exfat_date_time *tp)
> > +{
> > + time_t second = ts->tv_sec;
> > + time_t day, month, year;
> > + time_t ld; /* leap day */
>
> Question for other maintainers: Has kernel code already time_t defined
> as 64bit? Or it is still just 32bit and 32bit systems and some time64_t
> needs to be used? I remember that there was discussion about these
> problems, but do not know if it was changed/fixed or not... Just a
> pointer for possible Y2038 problem. As "ts" is of type timespec64, but
> "second" of type time_t.
>
> > +
> > + /* Treats as local time with proper time */
> > + second -= sys_tz.tz_minuteswest * SECS_PER_MIN;
> > + tp->timezone.valid = 1;
> > + tp->timezone.off = TIMEZONE_CUR_OFFSET();
> > +
> > + /* Jan 1 GMT 00:00:00 1980. But what about another time zone? */
> > + if (second < UNIX_SECS_1980) {
> > + tp->second = 0;
> > + tp->minute = 0;
> > + tp->hour = 0;
> > + tp->day = 1;
> > + tp->month = 1;
> > + tp->year = 0;
> > + return;
> > + }
> > +
> > + if (second >= UNIX_SECS_2108) {
>
> Hello, this code cause compile errors on 32bit systems as UNIX_SECS_2108
> macro is not defined when BITS_PER_LONG == 32.
>
> Value 4354819200 really cannot fit into 32bit signed integer, so you
> should use 64bit signed integer. I would suggest to define this macro
> value via LL not just L suffix (and it would work on both 32 and 64bit)
>
> #define UNIX_SECS_2108 4354819200LL
>
> > + tp->second = 59;
> > + tp->minute = 59;
> > + tp->hour = 23;
> > + tp->day = 31;
> > + tp->month = 12;
> > + tp->year = 127;
> > + return;
> > + }
>
> --
> Pali Roh?r
> [email protected]
---end quoted text---
> On Wed, Jan 8, 2020 at 7:03 PM Christoph Hellwig <[email protected]> wrote:
> >
> > Arnd, can you review the exfat time handling, especially vs y2038
> > related issues?
>
> Sure, thanks for adding me to the loop
>
> > On Thu, Jan 02, 2020 at 10:19:02AM +0100, Pali Rohár wrote:
> > > On Thursday 02 January 2020 16:20:32 Namjae Jeon wrote:
> > > > +#define TIMEZONE_CUR_OFFSET() ((sys_tz.tz_minuteswest / (-15))
> & 0x7F)
> > > > +/* Convert linear UNIX date to a FAT time/date pair. */ void
> > > > +exfat_time_unix2fat(struct exfat_sb_info *sbi, struct timespec64
> *ts,
> > > > + struct exfat_date_time *tp) {
> > > > + time_t second = ts->tv_sec;
> > > > + time_t day, month, year;
> > > > + time_t ld; /* leap day */
> > >
> > > Question for other maintainers: Has kernel code already time_t
> > > defined as 64bit? Or it is still just 32bit and 32bit systems and
> > > some time64_t needs to be used? I remember that there was discussion
> > > about these problems, but do not know if it was changed/fixed or
> > > not... Just a pointer for possible Y2038 problem. As "ts" is of type
> > > timespec64, but "second" of type time_t.
>
> I am actually very close to sending the patches to remove the time_t
> definition from the kernel, at least in yesterday's version there were no
> users.
>
> exfat_time_unix2fat() seems to be a copy of the old fat_time_unix2fat()
> that we fixed a while ago, please have a look at that implementation based
> on time64_to_tm(), which avoids time_t.
Okay, Pali reported it and suggested to check your patch in staging/exfat.
I will fix it on v10.
Thanks for your review!
>
> Arnd