Subject: | |
From: | |
Reply To: | |
Date: | Tue, 30 Jul 1996 07:00:27 PDT |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
Stan writes:
>Paul writes:
>> A minor detail here: label table can have multiple extent blocks (each
>> extent block holds upto 20 extents). Giving the fact that each label table
>interesting...I know there *was* a limitation of 20 blocks, possibly due to
>early coding time limits?, on the label table...even though the data structures
>would have allowed a chain of arbitrary length. I'd been told it was
>due to limitations on the code "mapping" the Label Table to a virtual
>address range during bootup, maybe.
This is the magic that Paul alludes to. There is an entry in the label table
for every file on a volume including the label table itself. In essence, the
system doesn't know where a label table is until that label table is mapped
into memory as part of mounting a volume. A tricky business indeed.
>> extent must be exactly .75 MB, just 20 extents would severely limit the
>> number of files we can have on a disk.
>That may be theory...but I've certainly seen systems whose label
>table extents differ dignificantly from .75MB. I just checked 2 user's
>machines, and found extents only of about 3 MB. The first extent in each
>case was 769 pages, and the others 768 pages (one page = 4,096 bytes).
>BTW, I can't tell how many extents my file has now...I'm emailing from home,
>and the machine running the "add an extent" utility seems to have died :(
>I'll report why, when I find out.
Joe Searle [log in to unmask]
|
|
|