List for cgit developers and users
 help / color / mirror / Atom feed
From: John Keeping <john@keeping.me.uk>
To: Hristo Venev <hristo@venev.name>
Cc: cgit@lists.zx2c4.com
Subject: Re: [PATCH] cache: Tolerate short writes in print_slot
Date: Thu, 7 Oct 2021 10:35:49 +0100	[thread overview]
Message-ID: <YV6/dYmy+OsWuTMQ@keeping.me.uk> (raw)
In-Reply-To: <20210910141841.2092532-1-hristo@venev.name>

On Fri, Sep 10, 2021 at 05:18:41PM +0300, Hristo Venev wrote:
> sendfile() can return after a short read/write, so we may need to call
> it more than once. Furthermore, not all files support sendfile(), so we
> may need to fall back to read/write.

Have you seen these errors in practice, or is this just theoretical?

In recent (since v2.6.33) versions of Linux, all files should support
sendfile(), especially since we expect out_fd to be a socket or pipe.

> On the read/write path, use write_in_full which deals with short writes.
> 
> Signed-off-by: Hristo Venev <hristo@venev.name>
> ---
>  cache.c | 46 ++++++++++++++++++++++++----------------------
>  1 file changed, 24 insertions(+), 22 deletions(-)
> 
> diff --git a/cache.c b/cache.c
> index 55199e8..85cfbd9 100644
> --- a/cache.c
> +++ b/cache.c
> @@ -85,40 +85,42 @@ static int close_slot(struct cache_slot *slot)
>  /* Print the content of the active cache slot (but skip the key). */
>  static int print_slot(struct cache_slot *slot)
>  {
> -#ifdef HAVE_LINUX_SENDFILE
> -	off_t start_off;
> -	int ret;
> +	off_t off;
> +	ssize_t i;
> +
> +	off = slot->keylen + 1;
>  
> -	start_off = slot->keylen + 1;
> +#ifdef HAVE_LINUX_SENDFILE
> +	off_t size;

decl-after-stmt if HAVE_LINUX_SENDFILE is set.

> +	size = slot->cache_st.st_size;
>  
>  	do {
> -		ret = sendfile(STDOUT_FILENO, slot->cache_fd, &start_off,
> -				slot->cache_st.st_size - start_off);
> -		if (ret < 0) {
> +		i = sendfile(STDOUT_FILENO, slot->cache_fd, &off, size - off);

Why is ret renamed?  i is normally a loop index variable, using it for
the return value here is strange, please stick with "ret".

> +		if (i < 0) {
>  			if (errno == EAGAIN || errno == EINTR)
>  				continue;
> +			/* Fall back to read/write on EINVAL */
> +			if (errno == EINVAL)
> +				break;
>  			return errno;
>  		}
> -		return 0;
> +		if (off == size)
> +			return 0;
>  	} while (1);
> -#else
> -	ssize_t i, j;
> +#endif
>  
> -	i = lseek(slot->cache_fd, slot->keylen + 1, SEEK_SET);
> -	if (i != slot->keylen + 1)
> +	if (lseek(slot->cache_fd, off, SEEK_SET) != off)
>  		return errno;
>  
>  	do {
> -		i = j = xread(slot->cache_fd, slot->buf, sizeof(slot->buf));
> -		if (i > 0)
> -			j = xwrite(STDOUT_FILENO, slot->buf, i);
> -	} while (i > 0 && j == i);
> -
> -	if (i < 0 || j != i)
> -		return errno;
> -	else
> -		return 0;
> -#endif
> +		i = xread(slot->cache_fd, slot->buf, sizeof(slot->buf));
> +		if (i < 0)
> +			return errno;
> +		if (i == 0)
> +			return 0;
> +		if (write_in_full(STDOUT_FILENO, slot->buf, i) < 0)
> +			return errno;
> +	} while (1);
>  }
>  
>  /* Check if the slot has expired */
> -- 
> 2.31.1
> 

  reply	other threads:[~2021-10-07  9:36 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-09-10 14:18 Hristo Venev
2021-10-07  9:35 ` John Keeping [this message]
2021-10-07 15:44   ` Hristo Venev
2022-05-07 21:28     ` Jason A. Donenfeld
2022-05-07 21:32       ` Hristo Venev
2022-05-07 21:38         ` Jason A. Donenfeld

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=YV6/dYmy+OsWuTMQ@keeping.me.uk \
    --to=john@keeping.me.uk \
    --cc=cgit@lists.zx2c4.com \
    --cc=hristo@venev.name \
    /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).