Re: [RFC v6 05/23] linux: Use 32-bit time_t for rusage

[Alistair Francis <alistair23 at gmail dot com>]

On Tue, Jan 14, 2020 at 9:58 PM Lukasz Majewski <lukma@denx.de> wrote:
>
> Hi Alistair,
>
> > The Linux kernel expects rusage to use a 32-bit time_t, even on archs
> > with a 64-bit time_t (like RV32). To address this let's convert
> > rusage to/from 32-bit and 64-bit to ensure the kernel always gets
> > a 32-bit time_t.
> >
> > This means that all 32-bit architectures with a 64-bit time_t will be
> > able to use this generic implementation.
> >
> > This code is bsaed on similar code in alpha, but adjusted to pass the
>                ^^^^^ - based?
>
> > 32-bit time_t to the kernel.
> >
> > We can't directly call __wait4 as it expects a structrusage but we
>                                                  ^^^^^^^^^^^^ - struct
>                                                  rusage
>
> > have to pass in and use a struct rusage32. The same appies to
> > __getrusage. ---
> >  .../linux/generic/wordsize-32/getrusage.c     | 39 +++++++++
> >  .../linux/generic/wordsize-32/tv32-compat.h   | 47 +++++++++++
> >  .../sysv/linux/generic/wordsize-32/wait4.c    | 79
> > +++++++++++++++++++ 3 files changed, 165 insertions(+)
> >  create mode 100644
> > sysdeps/unix/sysv/linux/generic/wordsize-32/getrusage.c create mode
> > 100644 sysdeps/unix/sysv/linux/generic/wordsize-32/wait4.c
> >
> > diff --git a/sysdeps/unix/sysv/linux/generic/wordsize-32/getrusage.c
> > b/sysdeps/unix/sysv/linux/generic/wordsize-32/getrusage.c new file
> > mode 100644 index 0000000000..2ada94a954
> > --- /dev/null
> > +++ b/sysdeps/unix/sysv/linux/generic/wordsize-32/getrusage.c
> > @@ -0,0 +1,39 @@
> > +/* utimes -- change file timestamps.  Linux/RV32/tv32 version.
>
> Please update the in-comment header.

Fixed!

>
> > +   Copyright (C) 2020 Free Software Foundation, Inc.
> > +   This file is part of the GNU C Library.
> > +
> > +   The GNU C Library is free software; you can redistribute it and/or
> > +   modify it under the terms of the GNU Lesser General Public
> > +   License as published by the Free Software Foundation; either
> > +   version 2.1 of the License, or (at your option) any later version.
> > +
> > +   The GNU C Library is distributed in the hope that it will be
> > useful,
> > +   but WITHOUT ANY WARRANTY; without even the implied warranty of
> > +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> > +   Lesser General Public License for more details.
> > +
> > +   You should have received a copy of the GNU Lesser General Public
> > +   License along with the GNU C Library; if not, see
> > +   <http://www.gnu.org/licenses/>.  */
> > +
> > +#include <sys/time.h>
> > +#include <sys/resource.h>
> > +#include <sysdep.h>
> > +#include <tv32-compat.h>
> > +
> > +int
> > +__getrusage (int who, struct rusage *usage)
> > +{
> > +#if __TIMESIZE == 64
> > +  struct rusage32 usage32;
> > +  if (INLINE_SYSCALL_CALL (getrusage, who, &usage32) == -1)
> > +    return -1;
> > +
> > +  rusage32_to_rusage64 (&usage32, usage);
> > +  return 0;
> > +#else
> > +  return INLINE_SYSCALL_CALL (getrusage, who, usage);
> > +#endif
> > +}
>
> This is the same issue as pointed out with __setitimer support.
>
> Following the conversion convention (as in e.g. clock_settime.c) would
> allow Y2038 safe 32 bit ports benefit from this patch.

I commented on the other patch, I'm not clear what you mean here.

>
> > +
> > +strong_alias (__getrusage, getrusage)
> > diff --git
> > a/sysdeps/unix/sysv/linux/generic/wordsize-32/tv32-compat.h
> > b/sysdeps/unix/sysv/linux/generic/wordsize-32/tv32-compat.h index
> > aa960a6632..8a4bd15002 100644 ---
> > a/sysdeps/unix/sysv/linux/generic/wordsize-32/tv32-compat.h +++
> > b/sysdeps/unix/sysv/linux/generic/wordsize-32/tv32-compat.h @@ -24,6
> > +24,7 @@ #include <bits/types.h> #include <bits/types/time_t.h>
> >  #include <bits/types/struct_timeval.h>
> > +#include <bits/types/struct_rusage.h>
> >
> >  /* Structures containing 'struct timeval' with 32-bit time_t.  */
> >  struct itimerval32
> > @@ -32,4 +33,50 @@ struct itimerval32
> >    struct __timeval_long it_value;
> >  };
> >
> > +struct rusage32
>
> struct __rusage32 ?

Good idea, updated

>
> > +{
> > +  struct __timeval_long ru_utime;    /* user time used */
> > +  struct __timeval_long ru_stime;    /* system time used */
>
>             As pointed out in the other comment - maybe struct
>             __timeval32 would be more appropriate?
>
>         This is the same convention as we already have with struct
>         __timespec64 and __itimerspec64

Yeah, I think you are probably right. I'll wait and see what others
think of the other thread though.

>
> > +  long ru_maxrss;            /* maximum resident set size */
> > +  long ru_ixrss;             /* integral shared memory size */
> > +  long ru_idrss;             /* integral unshared data size */
> > +  long ru_isrss;             /* integral unshared stack size */
> > +  long ru_minflt;            /* page reclaims */
> > +  long ru_majflt;            /* page faults */
> > +  long ru_nswap;             /* swaps */
> > +  long ru_inblock;           /* block input operations */
> > +  long ru_oublock;           /* block output operations */
> > +  long ru_msgsnd;            /* messages sent */
> > +  long ru_msgrcv;            /* messages received */
> > +  long ru_nsignals;          /* signals received */
> > +  long ru_nvcsw;             /* voluntary context switches */
> > +  long ru_nivcsw;            /* involuntary " */
> > +};
> > +
> > +static inline void
> > +rusage32_to_rusage64 (const struct rusage32 *restrict r32,
> > +                      struct rusage *restrict r64)
> > +{
> > +  /* Make sure the entire output structure is cleared, including
> > +     padding and reserved fields.  */
> > +  memset (r64, 0, sizeof *r64);
> > +
> > +  r64->ru_utime    = valid_timeval_long_to_timeval64 (r32->ru_utime);
> > +  r64->ru_stime    = valid_timeval_long_to_timeval64 (r32->ru_stime);
> > +  r64->ru_maxrss   = r32->ru_maxrss;
> > +  r64->ru_ixrss    = r32->ru_ixrss;
> > +  r64->ru_idrss    = r32->ru_idrss;
> > +  r64->ru_isrss    = r32->ru_isrss;
> > +  r64->ru_minflt   = r32->ru_minflt;
> > +  r64->ru_majflt   = r32->ru_majflt;
> > +  r64->ru_nswap    = r32->ru_nswap;
> > +  r64->ru_inblock  = r32->ru_inblock;
> > +  r64->ru_oublock  = r32->ru_oublock;
> > +  r64->ru_msgsnd   = r32->ru_msgsnd;
> > +  r64->ru_msgrcv   = r32->ru_msgrcv;
> > +  r64->ru_nsignals = r32->ru_nsignals;
> > +  r64->ru_nvcsw    = r32->ru_nvcsw;
> > +  r64->ru_nivcsw   = r32->ru_nivcsw;
> > +}
> > +
> >  #endif /* tv32-compat.h */
> > diff --git a/sysdeps/unix/sysv/linux/generic/wordsize-32/wait4.c
> > b/sysdeps/unix/sysv/linux/generic/wordsize-32/wait4.c new file mode
> > 100644 index 0000000000..a4a81a0587
> > --- /dev/null
> > +++ b/sysdeps/unix/sysv/linux/generic/wordsize-32/wait4.c
> > @@ -0,0 +1,79 @@
> > +/* wait4 -- wait for process to change state.  Linux/RV32/tv32
> > version.
> > +   Copyright (C) 2020 Free Software Foundation, Inc.
> > +   This file is part of the GNU C Library.
> > +
> > +   The GNU C Library is free software; you can redistribute it and/or
> > +   modify it under the terms of the GNU Lesser General Public
> > +   License as published by the Free Software Foundation; either
> > +   version 2.1 of the License, or (at your option) any later version.
> > +
> > +   The GNU C Library is distributed in the hope that it will be
> > useful,
> > +   but WITHOUT ANY WARRANTY; without even the implied warranty of
> > +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> > +   Lesser General Public License for more details.
> > +
> > +   You should have received a copy of the GNU Lesser General Public
> > +   License along with the GNU C Library; if not, see
> > +   <http://www.gnu.org/licenses/>.  */
> > +
> > +#include <sys/wait.h>
> > +#include <sys/resource.h>
> > +#include <sysdep-cancel.h>
> > +#include <tv32-compat.h>
> > +
> > +pid_t
> > +__wait4 (pid_t pid, int *stat_loc, int options, struct rusage *usage)
> > +{
> > +  struct rusage32 usage32;
> > +  idtype_t idtype = P_PID;
> > +
> > +  if (pid < -1)
> > +    {
> > +      idtype = P_PGID;
> > +      pid *= -1;
> > +    }
> > +  else if (pid == -1)
> > +    idtype = P_ALL;
> > +  else if (pid == 0)
> > +    idtype = P_PGID;
> > +
> > +  options |= WEXITED;
> > +
> > +  siginfo_t infop;
> > +  if (SYSCALL_CANCEL (waitid, idtype, pid, &infop, options,
> > &usage32) < 0)
> > +    return -1;
> > +
> > +  if (stat_loc)
> > +    {
> > +      switch (infop.si_code)
> > +        {
> > +        case CLD_EXITED:
> > +          *stat_loc = W_EXITCODE (infop.si_status, 0);
> > +          break;
> > +        case CLD_DUMPED:
> > +          *stat_loc = WCOREFLAG | infop.si_status;
> > +    break;
> > +        case CLD_KILLED:
> > +          *stat_loc = infop.si_status;
> > +          break;
> > +        case CLD_TRAPPED:
> > +        case CLD_STOPPED:
> > +          *stat_loc = W_STOPCODE (infop.si_status);
> > +          break;
> > +        case CLD_CONTINUED:
> > +          *stat_loc = __W_CONTINUED;
> > +          break;
> > +  default:
> > +    *stat_loc = 0;
> > +    break;
> > +        }
> > +    }
> > +
> > +  if (usage != NULL)
> > +    rusage32_to_rusage64 (&usage32, usage);
> > +
> > +  return infop.si_pid;
> > +}
> > +
> > +libc_hidden_def (__wait4);
> > +weak_alias (__wait4, wait4)
>
> This wait4 implementation (as it doesn't use #define __TIMESIZE == 64)
> could be used by all 32 bit ports. Am I right?

Yes, all of these changes apply to all 32-bit ports.

Alistair

>
>
> Best regards,
>
> Lukasz Majewski
>
> --
>
> DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
> HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
> Phone: (+49)-8142-66989-59 Fax: (+49)-8142-66989-80 Email: lukma@denx.de