Re: Request for Comments: libarchive, bsdtar

From: Don Lewis (truckman_at_FreeBSD.org)
Date: 01/17/04

  • Next message: Don Lewis: "Re: Signal delivery to kernel threads/processes?" 
    Date: Fri, 16 Jan 2004 15:33:10 -0800 (PST)
To: kientzle@acm.org

    On 16 Jan, Tim Kientzle wrote:
    > On 15 Jan, Tim Robbins wrote:
    >>- Padding gzip'd tar archives (with bsdtar czf) causes [various problems]
    >
    > I've updated libarchive so that:
    > * last-block padding is set with a separate API call that can
    > be invoked at anytime before the archive is closed
    > (In particular, it can be invoked from within the client open
    > callback.)
    > * If it is not set manually, then the default behavior is:
    > = uncompressed data within a gzip/bzip2 compressed stream
    > is always padded
    > = if archive_write_open_file is used, then the last block
    > is padded if the output is stdout or a character or block
    > device, otherwise the last block is not padded
    > = if archive_write_open_file is not used, then the
    > "default default" behavior is for the last block
    > to not be padded. This may change.
    >
    > This appears to match the behavior of gtar.
    > I've updated bsdtar to simply use the library defaults.
    >
    > Don Lewis wrote:
    >> I would prefer to have explicit control of this behavior.
    >> BTW, ... an enhanced version of dd ... had the useful option "mobs", which was
    >> used to specify a minimum output block size.
    >
    > libarchive now has an API function: archive_write_set_bytes_in_last_block
    >
    > Unfortunately, the name is a bit misleading; suggestions appreciated.
    >
    > As a special case, if the argument to this function is zero, the
    > last block will be padded to the full block size. Otherwise, the
    > last block will be padded to a multiple of the indicated value.
    > For example, if you specify 1024, and the block was 1025 bytes, it will
    > get padded to 2048. If you specify 1, no padding will be added.
    > However, in no case will the last block be padded to be larger than the
    > archive block size (as set with archive_write_set_bytes_per_block).
    >
    > As described above, the archive_write_open_file function will
    > set this for you only if you have not already invoked it manually.
    > If you use the low-level archive_write_open, then you're on your own.
    >
    > Does this provide the "explicit control" you were looking for?

    Sounds good to me.
    _______________________________________________
    freebsd-arch@freebsd.org mailing list
    http://lists.freebsd.org/mailman/listinfo/freebsd-arch
    To unsubscribe, send any mail to "freebsd-arch-unsubscribe@freebsd.org"

  • Next message: Don Lewis: "Re: Signal delivery to kernel threads/processes?"