1. Patchew
  2. QEMU
  3. View series

[PATCH] system/rtc: Fix a possible year-2038 integer overflow problem

Thomas Huth posted 1 patch 1 week, 5 days ago
Patches applied successfully (tree, apply log)
git fetch https://github.com/patchew-project/qemu tags/patchew/20260518160255.1844308-1-thuth@redhat.com
There is a newer version of this series
system/rtc.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
[PATCH] system/rtc: Fix a possible year-2038 integer overflow problem
Posted by Thomas Huth 1 week, 5 days ago 
From: Thomas Huth <thuth@redhat.com>

rtc_realtime_clock_offset is initialized with:

  rtc_realtime_clock_offset = qemu_clock_get_ms(QEMU_CLOCK_REALTIME) / 1000;

And QEMU_CLOCK_REALTIME might be based on gettimeofday() in certain
cases (see get_clock_realtime() in include/qemu/timer.h). So this
counter will exceed 32 bits in the year 2038, thus we should not
store this value in a normal integer variable. Change it to an int64_t
to fix the problem.
And while we're at it, also adjust the nearby rtc_host_datetime_offset
variable to be on the safe side in the related code.

Signed-off-by: Thomas Huth <thuth@redhat.com>
---
 system/rtc.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/system/rtc.c b/system/rtc.c
index f13dd392a76..83b28088054 100644
--- a/system/rtc.c
+++ b/system/rtc.c
@@ -41,8 +41,8 @@ static enum {
     RTC_BASE_DATETIME,
 } rtc_base_type = RTC_BASE_UTC;
 static time_t rtc_ref_start_datetime;
-static int rtc_realtime_clock_offset; /* used only with QEMU_CLOCK_REALTIME */
-static int rtc_host_datetime_offset = -1; /* valid & used only with
+static int64_t rtc_realtime_clock_offset; /* used only with QEMU_CLOCK_REALTIME */
+static int64_t rtc_host_datetime_offset = -1; /* valid & used only with
                                              RTC_BASE_DATETIME */
 QEMUClockType rtc_clock;
 /***********************************************************/
-- 
2.54.0
Re: [PATCH] system/rtc: Fix a possible year-2038 integer overflow problem
Posted by Laurent Vivier 1 week, 5 days ago 
Le 18/05/2026 à 18:02, Thomas Huth a écrit :
> From: Thomas Huth <thuth@redhat.com>
> 
> rtc_realtime_clock_offset is initialized with:
> 
>    rtc_realtime_clock_offset = qemu_clock_get_ms(QEMU_CLOCK_REALTIME) / 1000;
> 
> And QEMU_CLOCK_REALTIME might be based on gettimeofday() in certain
> cases (see get_clock_realtime() in include/qemu/timer.h). So this
> counter will exceed 32 bits in the year 2038, thus we should not
> store this value in a normal integer variable. Change it to an int64_t
> to fix the problem.
> And while we're at it, also adjust the nearby rtc_host_datetime_offset
> variable to be on the safe side in the related code.
> 
> Signed-off-by: Thomas Huth <thuth@redhat.com>
> ---
>   system/rtc.c | 4 ++--
>   1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/system/rtc.c b/system/rtc.c
> index f13dd392a76..83b28088054 100644
> --- a/system/rtc.c
> +++ b/system/rtc.c
> @@ -41,8 +41,8 @@ static enum {
>       RTC_BASE_DATETIME,
>   } rtc_base_type = RTC_BASE_UTC;
>   static time_t rtc_ref_start_datetime;
> -static int rtc_realtime_clock_offset; /* used only with QEMU_CLOCK_REALTIME */
> -static int rtc_host_datetime_offset = -1; /* valid & used only with
> +static int64_t rtc_realtime_clock_offset; /* used only with QEMU_CLOCK_REALTIME */
> +static int64_t rtc_host_datetime_offset = -1; /* valid & used only with
>                                                RTC_BASE_DATETIME */
>   QEMUClockType rtc_clock;
>   /***********************************************************/

Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Re: [PATCH] system/rtc: Fix a possible year-2038 integer overflow problem
Posted by Daniel P. Berrangé 1 week, 5 days ago 
On Mon, May 18, 2026 at 06:02:55PM +0200, Thomas Huth wrote:
> From: Thomas Huth <thuth@redhat.com>
> 
> rtc_realtime_clock_offset is initialized with:
> 
>   rtc_realtime_clock_offset = qemu_clock_get_ms(QEMU_CLOCK_REALTIME) / 1000;
> 
> And QEMU_CLOCK_REALTIME might be based on gettimeofday() in certain
> cases (see get_clock_realtime() in include/qemu/timer.h). So this
> counter will exceed 32 bits in the year 2038, thus we should not
> store this value in a normal integer variable. Change it to an int64_t
> to fix the problem.
> And while we're at it, also adjust the nearby rtc_host_datetime_offset
> variable to be on the safe side in the related code.
> 
> Signed-off-by: Thomas Huth <thuth@redhat.com>
> ---
>  system/rtc.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)

Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>


With regards,
Daniel
-- 
|: https://berrange.com       ~~        https://hachyderm.io/@berrange :|
|: https://libvirt.org          ~~          https://entangle-photo.org :|
|: https://pixelfed.art/berrange   ~~    https://fstop138.berrange.com :|
Re: [PATCH] system/rtc: Fix a possible year-2038 integer overflow problem
Posted by Daniel P. Berrangé 1 week, 5 days ago 
On Mon, May 18, 2026 at 05:10:36PM +0100, Daniel P. Berrangé wrote:
> On Mon, May 18, 2026 at 06:02:55PM +0200, Thomas Huth wrote:
> > From: Thomas Huth <thuth@redhat.com>
> > 
> > rtc_realtime_clock_offset is initialized with:
> > 
> >   rtc_realtime_clock_offset = qemu_clock_get_ms(QEMU_CLOCK_REALTIME) / 1000;
> > 
> > And QEMU_CLOCK_REALTIME might be based on gettimeofday() in certain
> > cases (see get_clock_realtime() in include/qemu/timer.h). So this
> > counter will exceed 32 bits in the year 2038, thus we should not
> > store this value in a normal integer variable. Change it to an int64_t
> > to fix the problem.
> > And while we're at it, also adjust the nearby rtc_host_datetime_offset
> > variable to be on the safe side in the related code.
> > 
> > Signed-off-by: Thomas Huth <thuth@redhat.com>
> > ---
> >  system/rtc.c | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

Actually on second reading I think this patch is not quite right.
The code which uses the value is:

    time_t value = qemu_clock_get_ms(clock) / 1000;
    switch (clock) {
    case QEMU_CLOCK_REALTIME:
        value -= rtc_realtime_clock_offset;


On 64-bit platforms 'int' was 64-bit already and so was time_t so we
have no bug to fix.

On 32-bit platforms the patch fixes 'int' to 'int64' which is fine
but that int64 value is then subtracted from time_t which is usually
going to be 32-bit again, unless the OS was built with 64-bit time_t
on 32-bit which almost no one does.

IOW we only platform this fixes is 32-bit OS with 64-bit time_t.

We can't do any better than that as the caller code paths need a
time_t to pass into gmtime/localtime.

So IMHO we should just declare it as time_t. 

With regards,
Daniel
-- 
|: https://berrange.com       ~~        https://hachyderm.io/@berrange :|
|: https://libvirt.org          ~~          https://entangle-photo.org :|
|: https://pixelfed.art/berrange   ~~    https://fstop138.berrange.com :|
Re: [PATCH] system/rtc: Fix a possible year-2038 integer overflow problem
Posted by Philippe Mathieu-Daudé 1 week, 4 days ago 
On 18/5/26 18:18, Daniel P. Berrangé wrote:
> On Mon, May 18, 2026 at 05:10:36PM +0100, Daniel P. Berrangé wrote:
>> On Mon, May 18, 2026 at 06:02:55PM +0200, Thomas Huth wrote:
>>> From: Thomas Huth <thuth@redhat.com>
>>>
>>> rtc_realtime_clock_offset is initialized with:
>>>
>>>    rtc_realtime_clock_offset = qemu_clock_get_ms(QEMU_CLOCK_REALTIME) / 1000;
>>>
>>> And QEMU_CLOCK_REALTIME might be based on gettimeofday() in certain
>>> cases (see get_clock_realtime() in include/qemu/timer.h). So this
>>> counter will exceed 32 bits in the year 2038, thus we should not
>>> store this value in a normal integer variable. Change it to an int64_t
>>> to fix the problem.
>>> And while we're at it, also adjust the nearby rtc_host_datetime_offset
>>> variable to be on the safe side in the related code.
>>>
>>> Signed-off-by: Thomas Huth <thuth@redhat.com>
>>> ---
>>>   system/rtc.c | 4 ++--
>>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
> 
> Actually on second reading I think this patch is not quite right.
> The code which uses the value is:
> 
>      time_t value = qemu_clock_get_ms(clock) / 1000;
>      switch (clock) {
>      case QEMU_CLOCK_REALTIME:
>          value -= rtc_realtime_clock_offset;
> 
> 
> On 64-bit platforms 'int' was 64-bit already and so was time_t so we
> have no bug to fix.
> 
> On 32-bit platforms the patch fixes 'int' to 'int64' which is fine
> but that int64 value is then subtracted from time_t which is usually
> going to be 32-bit again, unless the OS was built with 64-bit time_t
> on 32-bit which almost no one does.
> 
> IOW we only platform this fixes is 32-bit OS with 64-bit time_t.

"32-bit *host* OS".

Per ./configure script, we still allow:
- x86_64 x32 ABI      (CPU_CFLAGS="-mx32")
- sparc32             (CPU_CFLAGS="-m32 -mv8plus -mcpu=ultrasparc")
- s390                (CPU_CFLAGS="-m31")
otherwise falling back to TCI.

For sure, none of them are tested.

So, only TCI left?

> 
> We can't do any better than that as the caller code paths need a
> time_t to pass into gmtime/localtime.
> 
> So IMHO we should just declare it as time_t.
> 
> With regards,
> Daniel
Re: [PATCH] system/rtc: Fix a possible year-2038 integer overflow problem
Posted by Richard Henderson 1 week, 3 days ago 
On 5/19/26 01:24, Philippe Mathieu-Daudé wrote:
>> IOW we only platform this fixes is 32-bit OS with 64-bit time_t.
> 
> "32-bit *host* OS".
> 
> Per ./configure script, we still allow:
> - x86_64 x32 ABI      (CPU_CFLAGS="-mx32")
> - sparc32             (CPU_CFLAGS="-m32 -mv8plus -mcpu=ultrasparc")
> - s390                (CPU_CFLAGS="-m31")

All that is stuff that missed getting removed with the rest of 32-bit hosts.

> otherwise falling back to TCI.

Even TCI is 64-bit now.


r~
Re: [PATCH] system/rtc: Fix a possible year-2038 integer overflow problem
Posted by Thomas Huth 1 week, 4 days ago 
On 19/05/2026 10.24, Philippe Mathieu-Daudé wrote:
> On 18/5/26 18:18, Daniel P. Berrangé wrote:
>> On Mon, May 18, 2026 at 05:10:36PM +0100, Daniel P. Berrangé wrote:
>>> On Mon, May 18, 2026 at 06:02:55PM +0200, Thomas Huth wrote:
>>>> From: Thomas Huth <thuth@redhat.com>
>>>>
>>>> rtc_realtime_clock_offset is initialized with:
>>>>
>>>>    rtc_realtime_clock_offset = qemu_clock_get_ms(QEMU_CLOCK_REALTIME) / 
>>>> 1000;
>>>>
>>>> And QEMU_CLOCK_REALTIME might be based on gettimeofday() in certain
>>>> cases (see get_clock_realtime() in include/qemu/timer.h). So this
>>>> counter will exceed 32 bits in the year 2038, thus we should not
>>>> store this value in a normal integer variable. Change it to an int64_t
>>>> to fix the problem.
>>>> And while we're at it, also adjust the nearby rtc_host_datetime_offset
>>>> variable to be on the safe side in the related code.
>>>>
>>>> Signed-off-by: Thomas Huth <thuth@redhat.com>
>>>> ---
>>>>   system/rtc.c | 4 ++--
>>>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>>
>>> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
>>
>> Actually on second reading I think this patch is not quite right.
>> The code which uses the value is:
>>
>>      time_t value = qemu_clock_get_ms(clock) / 1000;
>>      switch (clock) {
>>      case QEMU_CLOCK_REALTIME:
>>          value -= rtc_realtime_clock_offset;
>>
>>
>> On 64-bit platforms 'int' was 64-bit already and so was time_t so we
>> have no bug to fix.
>>
>> On 32-bit platforms the patch fixes 'int' to 'int64' which is fine
>> but that int64 value is then subtracted from time_t which is usually
>> going to be 32-bit again, unless the OS was built with 64-bit time_t
>> on 32-bit which almost no one does.
>>
>> IOW we only platform this fixes is 32-bit OS with 64-bit time_t.
> 
> "32-bit *host* OS".
> 
> Per ./configure script, we still allow:
> - x86_64 x32 ABI      (CPU_CFLAGS="-mx32")

We should maybe deprecate that x32 ABI ...

> - sparc32             (CPU_CFLAGS="-m32 -mv8plus -mcpu=ultrasparc")

... and that one as well?

> - s390                (CPU_CFLAGS="-m31")

Uh, we really still allow 31-bit s390 hosts here? I think this could be 
removed nowadays, also upstream kernel recently ditched the 31-bit userspace:

  https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=8e0b986c59c67e08ada646249f834655a9e6da16

Question is whether we can remove it immediately, or whether it needs to be 
deprecated first?

  Thomas


> otherwise falling back to TCI.
> 
> For sure, none of them are tested.
> 
> So, only TCI left?
> 
>>
>> We can't do any better than that as the caller code paths need a
>> time_t to pass into gmtime/localtime.
>>
>> So IMHO we should just declare it as time_t.
>>
>> With regards,
>> Daniel
>
Re: [PATCH] system/rtc: Fix a possible year-2038 integer overflow problem
Posted by Thomas Huth 1 week, 4 days ago 
On 19/05/2026 11.18, Thomas Huth wrote:
> On 19/05/2026 10.24, Philippe Mathieu-Daudé wrote:
>> On 18/5/26 18:18, Daniel P. Berrangé wrote:
>>> On Mon, May 18, 2026 at 05:10:36PM +0100, Daniel P. Berrangé wrote:
>>>> On Mon, May 18, 2026 at 06:02:55PM +0200, Thomas Huth wrote:
>>>>> From: Thomas Huth <thuth@redhat.com>
>>>>>
>>>>> rtc_realtime_clock_offset is initialized with:
>>>>>
>>>>>    rtc_realtime_clock_offset = qemu_clock_get_ms(QEMU_CLOCK_REALTIME) / 
>>>>> 1000;
>>>>>
>>>>> And QEMU_CLOCK_REALTIME might be based on gettimeofday() in certain
>>>>> cases (see get_clock_realtime() in include/qemu/timer.h). So this
>>>>> counter will exceed 32 bits in the year 2038, thus we should not
>>>>> store this value in a normal integer variable. Change it to an int64_t
>>>>> to fix the problem.
>>>>> And while we're at it, also adjust the nearby rtc_host_datetime_offset
>>>>> variable to be on the safe side in the related code.
>>>>>
>>>>> Signed-off-by: Thomas Huth <thuth@redhat.com>
>>>>> ---
>>>>>   system/rtc.c | 4 ++--
>>>>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>>>
>>>> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
>>>
>>> Actually on second reading I think this patch is not quite right.
>>> The code which uses the value is:
>>>
>>>      time_t value = qemu_clock_get_ms(clock) / 1000;
>>>      switch (clock) {
>>>      case QEMU_CLOCK_REALTIME:
>>>          value -= rtc_realtime_clock_offset;
>>>
>>>
>>> On 64-bit platforms 'int' was 64-bit already and so was time_t so we
>>> have no bug to fix.
>>>
>>> On 32-bit platforms the patch fixes 'int' to 'int64' which is fine
>>> but that int64 value is then subtracted from time_t which is usually
>>> going to be 32-bit again, unless the OS was built with 64-bit time_t
>>> on 32-bit which almost no one does.
>>>
>>> IOW we only platform this fixes is 32-bit OS with 64-bit time_t.
>>
>> "32-bit *host* OS".
>>
>> Per ./configure script, we still allow:
>> - x86_64 x32 ABI      (CPU_CFLAGS="-mx32")
> 
> We should maybe deprecate that x32 ABI ...
> 
>> - sparc32             (CPU_CFLAGS="-m32 -mv8plus -mcpu=ultrasparc")
> 
> ... and that one as well?
> 
>> - s390                (CPU_CFLAGS="-m31")
> 
> Uh, we really still allow 31-bit s390 hosts here? I think this could be 
> removed nowadays, also upstream kernel recently ditched the 31-bit userspace:
> 
>   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/ 
> commit/?id=8e0b986c59c67e08ada646249f834655a9e6da16
> 
> Question is whether we can remove it immediately, or whether it needs to be 
> deprecated first?

Ah, wait, we completely disallow disallow 32-bit hosts nowadays:

  https://gitlab.com/qemu-project/qemu/-/commit/372ec46b9f1215f48a4717f2b7ed969f65bfadc6

... so these checks in the configure script are just dead code nowadays that 
could be removed now? Or is there still some hidden magic here that depends 
on these lines?

  Thomas
Re: [PATCH] system/rtc: Fix a possible year-2038 integer overflow problem
Posted by Alex Bennée 1 week, 4 days ago 
Thomas Huth <thuth@redhat.com> writes:

> On 19/05/2026 11.18, Thomas Huth wrote:
>> On 19/05/2026 10.24, Philippe Mathieu-Daudé wrote:
>>> On 18/5/26 18:18, Daniel P. Berrangé wrote:
>>>> On Mon, May 18, 2026 at 05:10:36PM +0100, Daniel P. Berrangé wrote:
>>>>> On Mon, May 18, 2026 at 06:02:55PM +0200, Thomas Huth wrote:
>>>>>> From: Thomas Huth <thuth@redhat.com>
>>>>>>
>>>>>> rtc_realtime_clock_offset is initialized with:
>>>>>>
>>>>>>    rtc_realtime_clock_offset =
>>>>>> qemu_clock_get_ms(QEMU_CLOCK_REALTIME) / 1000;
>>>>>>
>>>>>> And QEMU_CLOCK_REALTIME might be based on gettimeofday() in certain
>>>>>> cases (see get_clock_realtime() in include/qemu/timer.h). So this
>>>>>> counter will exceed 32 bits in the year 2038, thus we should not
>>>>>> store this value in a normal integer variable. Change it to an int64_t
>>>>>> to fix the problem.
>>>>>> And while we're at it, also adjust the nearby rtc_host_datetime_offset
>>>>>> variable to be on the safe side in the related code.
>>>>>>
>>>>>> Signed-off-by: Thomas Huth <thuth@redhat.com>
>>>>>> ---
>>>>>>   system/rtc.c | 4 ++--
>>>>>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>>>>
>>>>> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
>>>>
>>>> Actually on second reading I think this patch is not quite right.
>>>> The code which uses the value is:
>>>>
>>>>      time_t value = qemu_clock_get_ms(clock) / 1000;
>>>>      switch (clock) {
>>>>      case QEMU_CLOCK_REALTIME:
>>>>          value -= rtc_realtime_clock_offset;
>>>>
>>>>
>>>> On 64-bit platforms 'int' was 64-bit already and so was time_t so we
>>>> have no bug to fix.
>>>>
>>>> On 32-bit platforms the patch fixes 'int' to 'int64' which is fine
>>>> but that int64 value is then subtracted from time_t which is usually
>>>> going to be 32-bit again, unless the OS was built with 64-bit time_t
>>>> on 32-bit which almost no one does.
>>>>
>>>> IOW we only platform this fixes is 32-bit OS with 64-bit time_t.
>>>
>>> "32-bit *host* OS".
>>>
>>> Per ./configure script, we still allow:
>>> - x86_64 x32 ABI      (CPU_CFLAGS="-mx32")
>> We should maybe deprecate that x32 ABI ...
>> 
>>> - sparc32             (CPU_CFLAGS="-m32 -mv8plus -mcpu=ultrasparc")
>> ... and that one as well?
>> 
>>> - s390                (CPU_CFLAGS="-m31")
>> Uh, we really still allow 31-bit s390 hosts here? I think this could
>> be removed nowadays, also upstream kernel recently ditched the
>> 31-bit userspace:
>>   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/
>> commit/?id=8e0b986c59c67e08ada646249f834655a9e6da16
>> Question is whether we can remove it immediately, or whether it
>> needs to be deprecated first?
>
> Ah, wait, we completely disallow disallow 32-bit hosts nowadays:
>
>  https://gitlab.com/qemu-project/qemu/-/commit/372ec46b9f1215f48a4717f2b7ed969f65bfadc6
>
> ... so these checks in the configure script are just dead code
> nowadays that could be removed now? Or is there still some hidden
> magic here that depends on these lines?

Didn't we re-enable 32 bit hosts for qemu-img and friends?
>
>  Thomas

-- 
Alex Bennée
Virtualisation Tech Lead @ Linaro
Re: [PATCH] system/rtc: Fix a possible year-2038 integer overflow problem
Posted by Peter Maydell 1 week, 4 days ago 
On Tue, 19 May 2026 at 10:27, Thomas Huth <thuth@redhat.com> wrote:
>
> On 19/05/2026 11.18, Thomas Huth wrote:
> > On 19/05/2026 10.24, Philippe Mathieu-Daudé wrote:
> >> "32-bit *host* OS".
> >>
> >> Per ./configure script, we still allow:
> >> - x86_64 x32 ABI      (CPU_CFLAGS="-mx32")
> >
> > We should maybe deprecate that x32 ABI ...
> >
> >> - sparc32             (CPU_CFLAGS="-m32 -mv8plus -mcpu=ultrasparc")
> >
> > ... and that one as well?
> >
> >> - s390                (CPU_CFLAGS="-m31")

> > Uh, we really still allow 31-bit s390 hosts here? I think this could be
> > removed nowadays, also upstream kernel recently ditched the 31-bit userspace:
> >
> >   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/
> > commit/?id=8e0b986c59c67e08ada646249f834655a9e6da16
> >
> > Question is whether we can remove it immediately, or whether it needs to be
> > deprecated first?
>
> Ah, wait, we completely disallow disallow 32-bit hosts nowadays:
>
>   https://gitlab.com/qemu-project/qemu/-/commit/372ec46b9f1215f48a4717f2b7ed969f65bfadc6

We relaxed that check before the 11.0 release in commit cf634dfcd8fc.
We require a 64-bit host for the emulator binaries but still permit
32-bit host builds for the tools.

thanks
-- PMM
Re: [PATCH] system/rtc: Fix a possible year-2038 integer overflow problem
Posted by Thomas Huth 1 week, 4 days ago 
On 19/05/2026 12.21, Peter Maydell wrote:
> On Tue, 19 May 2026 at 10:27, Thomas Huth <thuth@redhat.com> wrote:
>>
>> On 19/05/2026 11.18, Thomas Huth wrote:
>>> On 19/05/2026 10.24, Philippe Mathieu-Daudé wrote:
>>>> "32-bit *host* OS".
>>>>
>>>> Per ./configure script, we still allow:
>>>> - x86_64 x32 ABI      (CPU_CFLAGS="-mx32")
>>>
>>> We should maybe deprecate that x32 ABI ...
>>>
>>>> - sparc32             (CPU_CFLAGS="-m32 -mv8plus -mcpu=ultrasparc")
>>>
>>> ... and that one as well?
>>>
>>>> - s390                (CPU_CFLAGS="-m31")
> 
>>> Uh, we really still allow 31-bit s390 hosts here? I think this could be
>>> removed nowadays, also upstream kernel recently ditched the 31-bit userspace:
>>>
>>>    https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/
>>> commit/?id=8e0b986c59c67e08ada646249f834655a9e6da16
>>>
>>> Question is whether we can remove it immediately, or whether it needs to be
>>> deprecated first?
>>
>> Ah, wait, we completely disallow disallow 32-bit hosts nowadays:
>>
>>    https://gitlab.com/qemu-project/qemu/-/commit/372ec46b9f1215f48a4717f2b7ed969f65bfadc6
> 
> We relaxed that check before the 11.0 release in commit cf634dfcd8fc.
> We require a 64-bit host for the emulator binaries but still permit
> 32-bit host builds for the tools.

Ah, thanks, that was the part that I missed!

  Thomas
Re: [PATCH] system/rtc: Fix a possible year-2038 integer overflow problem
Posted by Philippe Mathieu-Daudé 1 week, 4 days ago 
On 19/5/26 11:26, Thomas Huth wrote:
> On 19/05/2026 11.18, Thomas Huth wrote:
>> On 19/05/2026 10.24, Philippe Mathieu-Daudé wrote:
>>> On 18/5/26 18:18, Daniel P. Berrangé wrote:
>>>> On Mon, May 18, 2026 at 05:10:36PM +0100, Daniel P. Berrangé wrote:
>>>>> On Mon, May 18, 2026 at 06:02:55PM +0200, Thomas Huth wrote:
>>>>>> From: Thomas Huth <thuth@redhat.com>
>>>>>>
>>>>>> rtc_realtime_clock_offset is initialized with:
>>>>>>
>>>>>>    rtc_realtime_clock_offset = 
>>>>>> qemu_clock_get_ms(QEMU_CLOCK_REALTIME) / 1000;
>>>>>>
>>>>>> And QEMU_CLOCK_REALTIME might be based on gettimeofday() in certain
>>>>>> cases (see get_clock_realtime() in include/qemu/timer.h). So this
>>>>>> counter will exceed 32 bits in the year 2038, thus we should not
>>>>>> store this value in a normal integer variable. Change it to an 
>>>>>> int64_t
>>>>>> to fix the problem.
>>>>>> And while we're at it, also adjust the nearby 
>>>>>> rtc_host_datetime_offset
>>>>>> variable to be on the safe side in the related code.
>>>>>>
>>>>>> Signed-off-by: Thomas Huth <thuth@redhat.com>
>>>>>> ---
>>>>>>   system/rtc.c | 4 ++--
>>>>>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>>>>
>>>>> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
>>>>
>>>> Actually on second reading I think this patch is not quite right.
>>>> The code which uses the value is:
>>>>
>>>>      time_t value = qemu_clock_get_ms(clock) / 1000;
>>>>      switch (clock) {
>>>>      case QEMU_CLOCK_REALTIME:
>>>>          value -= rtc_realtime_clock_offset;
>>>>
>>>>
>>>> On 64-bit platforms 'int' was 64-bit already and so was time_t so we
>>>> have no bug to fix.
>>>>
>>>> On 32-bit platforms the patch fixes 'int' to 'int64' which is fine
>>>> but that int64 value is then subtracted from time_t which is usually
>>>> going to be 32-bit again, unless the OS was built with 64-bit time_t
>>>> on 32-bit which almost no one does.
>>>>
>>>> IOW we only platform this fixes is 32-bit OS with 64-bit time_t.
>>>
>>> "32-bit *host* OS".
>>>
>>> Per ./configure script, we still allow:
>>> - x86_64 x32 ABI      (CPU_CFLAGS="-mx32")
>>
>> We should maybe deprecate that x32 ABI ...
>>
>>> - sparc32             (CPU_CFLAGS="-m32 -mv8plus -mcpu=ultrasparc")
>>
>> ... and that one as well?
>>
>>> - s390                (CPU_CFLAGS="-m31")
>>
>> Uh, we really still allow 31-bit s390 hosts here? I think this could 
>> be removed nowadays, also upstream kernel recently ditched the 31-bit 
>> userspace:
>>
>>   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/ 
>> commit/?id=8e0b986c59c67e08ada646249f834655a9e6da16
>>
>> Question is whether we can remove it immediately, or whether it needs 
>> to be deprecated first?
> 
> Ah, wait, we completely disallow disallow 32-bit hosts nowadays:
> 
>   https://gitlab.com/qemu-project/qemu/-/ 
> commit/372ec46b9f1215f48a4717f2b7ed969f65bfadc6
> 
> ... so these checks in the configure script are just dead code nowadays 
> that could be removed now? Or is there still some hidden magic here that 
> depends on these lines?

Ah think, I remember reviewing something like that but was looking in
./configure history and forgot to look at meson.build. So this is indeed
dead code.
Re: [PATCH] system/rtc: Fix a possible year-2038 integer overflow problem
Posted by Thomas Huth 1 week, 5 days ago 
On 18/05/2026 18.18, Daniel P. Berrangé wrote:
> On Mon, May 18, 2026 at 05:10:36PM +0100, Daniel P. Berrangé wrote:
>> On Mon, May 18, 2026 at 06:02:55PM +0200, Thomas Huth wrote:
>>> From: Thomas Huth <thuth@redhat.com>
>>>
>>> rtc_realtime_clock_offset is initialized with:
>>>
>>>    rtc_realtime_clock_offset = qemu_clock_get_ms(QEMU_CLOCK_REALTIME) / 1000;
>>>
>>> And QEMU_CLOCK_REALTIME might be based on gettimeofday() in certain
>>> cases (see get_clock_realtime() in include/qemu/timer.h). So this
>>> counter will exceed 32 bits in the year 2038, thus we should not
>>> store this value in a normal integer variable. Change it to an int64_t
>>> to fix the problem.
>>> And while we're at it, also adjust the nearby rtc_host_datetime_offset
>>> variable to be on the safe side in the related code.
>>>
>>> Signed-off-by: Thomas Huth <thuth@redhat.com>
>>> ---
>>>   system/rtc.c | 4 ++--
>>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
> 
> Actually on second reading I think this patch is not quite right.
> The code which uses the value is:
> 
>      time_t value = qemu_clock_get_ms(clock) / 1000;
>      switch (clock) {
>      case QEMU_CLOCK_REALTIME:
>          value -= rtc_realtime_clock_offset;
> 
> 
> On 64-bit platforms 'int' was 64-bit already and so was time_t so we
> have no bug to fix.

Uh, did you mix this up with "long" ?

> On 32-bit platforms the patch fixes 'int' to 'int64' which is fine
> but that int64 value is then subtracted from time_t which is usually
> going to be 32-bit again, unless the OS was built with 64-bit time_t
> on 32-bit which almost no one does.
> 
> IOW we only platform this fixes is 32-bit OS with 64-bit time_t.
> 
> We can't do any better than that as the caller code paths need a
> time_t to pass into gmtime/localtime.
> 
> So IMHO we should just declare it as time_t.

Fine for me, too. I'll send a v2...

  Thomas

Patchew 2.1-devel-ea86937

© 2016 - 2026 Red Hat, Inc.