qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Warner Losh <imp@bsdimp.com>
To: Richard Henderson <richard.henderson@linaro.org>
Cc: Peter Maydell <peter.maydell@linaro.org>,
	Stacey Son <sson@freebsd.org>, Gleb Popov <arrowd@freebsd.org>,
	Kyle Evans <kevans@freebsd.org>,
	QEMU Developers <qemu-devel@nongnu.org>,
	Konrad Witaszczyk <def@freebsd.org>,
	Jessica Clarke <jrtc27@freebsd.org>,
	Brad Smith <brad@comstyle.com>
Subject: Re: [PATCH 20/22] bsd-user/bsd-file.h: Add implementations for read,  pread, readv and preadv
Date: Tue, 1 Feb 2022 20:02:26 -0700	[thread overview]
Message-ID: <CANCZdfo3-APWNr+cu3P=uD7-gYJvDSFqHeE98dutN26ACSd7Bg@mail.gmail.com> (raw)
In-Reply-To: <9863981c-9040-6eff-27e6-6629d94069a5@linaro.org>

[-- Attachment #1: Type: text/plain, Size: 5160 bytes --]

On Tue, Feb 1, 2022 at 2:37 PM Richard Henderson <
richard.henderson@linaro.org> wrote:

> On 2/1/22 22:14, Warner Losh wrote:
> > Implement do_bsd_{read,pread,readv,preadv}. Connect them to the system
> > call table.
> >
> > Signed-off-by: Stacey Son <sson@FreeBSD.org>
> > Signed-off-by: Kyle Evans <kevans@FreeBSD.org>
> > Signed-off-by: Warner Losh <imp@bsdimp.com>
> > ---
> >   bsd-user/bsd-file.h           | 79 +++++++++++++++++++++++++++++++++++
> >   bsd-user/freebsd/os-syscall.c | 24 +++++++++++
> >   2 files changed, 103 insertions(+)
> >
> > diff --git a/bsd-user/bsd-file.h b/bsd-user/bsd-file.h
> > index 2f743db38e1..5934cbd5612 100644
> > --- a/bsd-user/bsd-file.h
> > +++ b/bsd-user/bsd-file.h
> > @@ -36,4 +36,83 @@ extern struct iovec *lock_iovec(int type, abi_ulong
> target_addr, int count,
> >   extern void unlock_iovec(struct iovec *vec, abi_ulong target_addr, int
> count,
> >           int copy);
> >
> > +ssize_t safe_read(int fd, void *buf, size_t nbytes);
> > +ssize_t safe_pread(int fd, void *buf, size_t nbytes, off_t offset);
> > +ssize_t safe_readv(int fd, const struct iovec *iov, int iovcnt);
> > +ssize_t safe_preadv(int fd, const struct iovec *iov, int iovcnt, off_t
> offset);
> > +
> > +/* read(2) */
> > +static inline abi_long do_bsd_read(abi_long arg1, abi_long arg2,
> abi_long arg3)
> > +{
> > +    abi_long ret;
> > +    void *p;
> > +
> > +    p = lock_user(VERIFY_WRITE, arg2, arg3, 0);
> > +    if (p == NULL) {
> > +        return -TARGET_EFAULT;
> > +    }
> > +    ret = get_errno(safe_read(arg1, p, arg3));
> > +    unlock_user(p, arg2, ret);
> > +
> > +    return ret;
> > +}
> > +
> > +/* pread(2) */
> > +static inline abi_long do_bsd_pread(void *cpu_env, abi_long arg1,
> > +    abi_long arg2, abi_long arg3, abi_long arg4, abi_long arg5,
> abi_long arg6)
> > +{
> > +    abi_long ret;
> > +    void *p;
> > +
> > +    p = lock_user(VERIFY_WRITE, arg2, arg3, 0);
> > +    if (p == NULL) {
> > +        return -TARGET_EFAULT;
> > +    }
> > +    if (regpairs_aligned(cpu_env) != 0) {
> > +        arg4 = arg5;
> > +        arg5 = arg6;
> > +    }
>
> This would be clearer if you had started labeling from arg0.
>

There's a number of other changes this would force, so I'll defer
it...

Warner


> But either way,
> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
>
>
> r~
>
> > +    ret = get_errno(safe_pread(arg1, p, arg3, target_arg64(arg4,
> arg5)));
> > +    unlock_user(p, arg2, ret);
> > +
> > +    return ret;
> > +}
> > +
> > +/* readv(2) */
> > +static inline abi_long do_bsd_readv(abi_long arg1, abi_long arg2,
> abi_long arg3)
> > +{
> > +    abi_long ret;
> > +    struct iovec *vec = lock_iovec(VERIFY_WRITE, arg2, arg3, 0);
> > +
> > +    if (vec != NULL) {
> > +        ret = get_errno(safe_readv(arg1, vec, arg3));
> > +        unlock_iovec(vec, arg2, arg3, 1);
> > +    } else {
> > +        ret = -host_to_target_errno(errno);
> > +    }
> > +
> > +    return ret;
> > +}
> > +
> > +/* preadv(2) */
> > +static inline abi_long do_bsd_preadv(void *cpu_env, abi_long arg1,
> > +    abi_long arg2, abi_long arg3, abi_long arg4, abi_long arg5,
> abi_long arg6)
> > +{
> > +    abi_long ret;
> > +    struct iovec *vec = lock_iovec(VERIFY_WRITE, arg2, arg3, 1);
> > +
> > +    if (vec != NULL) {
> > +        if (regpairs_aligned(cpu_env) != 0) {
> > +            arg4 = arg5;
> > +            arg5 = arg6;
> > +        }
> > +        ret = get_errno(safe_preadv(arg1, vec, arg3, target_arg64(arg4,
> arg5)));
> > +        unlock_iovec(vec, arg2, arg3, 0);
> > +    } else {
> > +        ret = -host_to_target_errno(errno);
> > +    }
> > +
> > +    return ret;
> > +}
> > +
> >   #endif /* !BSD_FILE_H_ */
> > diff --git a/bsd-user/freebsd/os-syscall.c
> b/bsd-user/freebsd/os-syscall.c
> > index fcfa6221182..dda79af53de 100644
> > --- a/bsd-user/freebsd/os-syscall.c
> > +++ b/bsd-user/freebsd/os-syscall.c
> > @@ -42,6 +42,14 @@
> >
> >   #include "bsd-file.h"
> >
> > +/* I/O */
> > +safe_syscall3(ssize_t, read, int, fd, void *, buf, size_t, nbytes);
> > +safe_syscall4(ssize_t, pread, int, fd, void *, buf, size_t, nbytes,
> off_t,
> > +    offset);
> > +safe_syscall3(ssize_t, readv, int, fd, const struct iovec *, iov, int,
> iovcnt);
> > +safe_syscall4(ssize_t, preadv, int, fd, const struct iovec *, iov, int,
> iovcnt,
> > +    off_t, offset);
> > +
> >   void target_set_brk(abi_ulong new_brk)
> >   {
> >   }
> > @@ -212,6 +220,22 @@ abi_long do_freebsd_syscall(void *cpu_env, int num,
> abi_long arg1,
> >       }
> >
> >       switch (num) {
> > +
> > +        /*
> > +         * File system calls.
> > +         */
> > +    case TARGET_FREEBSD_NR_read: /* read(2) */
> > +        ret = do_bsd_read(arg1, arg2, arg3);
> > +        break;
> > +
> > +    case TARGET_FREEBSD_NR_pread: /* pread(2) */
> > +        ret = do_bsd_pread(cpu_env, arg1, arg2, arg3, arg4, arg5, arg6);
> > +        break;
> > +
> > +    case TARGET_FREEBSD_NR_readv: /* readv(2) */
> > +        ret = do_bsd_readv(arg1, arg2, arg3);
> > +        break;
> > +
> >       default:
> >           gemu_log("qemu: unsupported syscall: %d\n", num);
> >           ret = -TARGET_ENOSYS;
>
>

[-- Attachment #2: Type: text/html, Size: 6948 bytes --]

  reply	other threads:[~2022-02-02  3:29 UTC|newest]

Thread overview: 67+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-02-01 11:14 [PATCH 00/22] bsd-user: Start upstreaming the system calls Warner Losh
2022-02-01 11:14 ` [PATCH 01/22] bsd-user/main.c: Drop syscall flavor arg -bsd Warner Losh
2022-02-01 15:54   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 02/22] bsd-user/x86_64/target_arch_cpu.h: int $80 never was a BSD system call on amd64 Warner Losh
2022-02-01 15:57   ` Richard Henderson
2022-02-01 16:19   ` Kyle Evans
2022-02-01 11:14 ` [PATCH 03/22] bsd-user/x86_64/target_arch_cpu.h: Remove openbsd syscall Warner Losh
2022-02-01 16:22   ` Kyle Evans
2022-02-01 23:13     ` Warner Losh
2022-02-01 17:27   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 04/22] bsd-user/i386/target_arch_cpu.h: " Warner Losh
2022-02-01 16:24   ` Kyle Evans
2022-02-01 17:29   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 05/22] bsd-user/arm/target_arch_cpu.h: Only support FreeBSD sys calls Warner Losh
2022-02-01 16:32   ` Kyle Evans
2022-02-01 23:19     ` Warner Losh
2022-02-01 17:31   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 06/22] bsd-user/arm/target_arch_thread.h: Assume a FreeBSD target Warner Losh
2022-02-01 16:33   ` Kyle Evans
2022-02-01 17:34   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 07/22] bsd-user/x86_64/target_arch_thread.h: " Warner Losh
2022-02-01 16:36   ` Kyle Evans
2022-02-01 17:34   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 08/22] bsd-user: Remove bsd_type Warner Losh
2022-02-01 16:34   ` Kyle Evans
2022-02-01 17:35   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 09/22] bsd-user/freebsd/os-syscall.c: Move syscall processing here Warner Losh
2022-02-01 16:18   ` Kyle Evans
2022-02-01 17:37   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 10/22] bsd-user: Move system call building to os-syscall.c Warner Losh
2022-02-01 17:38   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 11/22] bsd-user/sycall.c: Now obsolete, remove Warner Losh
2022-02-01 16:35   ` Kyle Evans
2022-02-01 17:38   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 12/22] bsd-user/freebsd/os-syscall.c: Add get_errno and host_to_target_errno Warner Losh
2022-02-01 16:35   ` Kyle Evans
2022-02-01 17:39   ` Richard Henderson
2022-02-01 23:27     ` Warner Losh
2022-02-01 11:14 ` [PATCH 13/22] bsd-user/bsd-file.h: Implementation details for the filesystem calls Warner Losh
2022-02-01 16:47   ` Kyle Evans
2022-02-01 17:43   ` Richard Henderson
2022-02-01 23:55     ` Warner Losh
2022-02-26 16:24     ` Warner Losh
2022-02-01 11:14 ` [PATCH 14/22] bsd-user/freebsd/os-syscall.c: lock_iovec Warner Losh
2022-02-01 16:46   ` Kyle Evans
2022-02-01 21:16   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 15/22] bsd-user/freebsd/os-syscall.c: unlock_iovec Warner Losh
2022-02-01 16:37   ` Kyle Evans
2022-02-01 21:29   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 16/22] bsd-user/freebsd/os-syscall.c: Tracing and error boilerplate Warner Losh
2022-02-01 21:31   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 17/22] bsd-user: introduce target.h Warner Losh
2022-02-01 16:39   ` Kyle Evans
2022-02-01 21:32   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 18/22] bsd-user: Define target_arg64 Warner Losh
2022-02-01 16:41   ` Kyle Evans
2022-02-01 21:33   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 19/22] bsd-user: Add safe system call macros Warner Losh
2022-02-01 21:33   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 20/22] bsd-user/bsd-file.h: Add implementations for read, pread, readv and preadv Warner Losh
2022-02-01 21:37   ` Richard Henderson
2022-02-02  3:02     ` Warner Losh [this message]
2022-02-01 11:14 ` [PATCH 21/22] bsd-user/bsd-file.h: Meat of the write system calls Warner Losh
2022-02-01 21:38   ` Richard Henderson
2022-02-01 11:14 ` [PATCH 22/22] bsd-user/freebsd/os-syscall.c: Implement exit Warner Losh
2022-02-01 16:48   ` Kyle Evans
2022-02-01 21:39   ` Richard Henderson

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CANCZdfo3-APWNr+cu3P=uD7-gYJvDSFqHeE98dutN26ACSd7Bg@mail.gmail.com' \
    --to=imp@bsdimp.com \
    --cc=arrowd@freebsd.org \
    --cc=brad@comstyle.com \
    --cc=def@freebsd.org \
    --cc=jrtc27@freebsd.org \
    --cc=kevans@freebsd.org \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-devel@nongnu.org \
    --cc=richard.henderson@linaro.org \
    --cc=sson@freebsd.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).