"Machida, Hiroyuki" <[EMAIL PROTECTED]> writes:
> Once concern about global URL in general, it tends to be occupied
> by specific pattern, like accesses from one process or to on dir.
> It prevents to realize locality.
>
> I think it's better to have limitations like;
> entries for same proc
Machida, Hiroyuki wrote:
OGAWA Hirofumi wrote:
"Machida, Hiroyuki" <[EMAIL PROTECTED]> writes:
Right, it looks like TLB, which holds cache "Physical addres"
correponding to "Logical address". In this case, PID and file name
to be looked up, perform role of "Logical address".
But, there i
OGAWA Hirofumi wrote:
"Machida, Hiroyuki" <[EMAIL PROTECTED]> writes:
Right, it looks like TLB, which holds cache "Physical addres"
correponding to "Logical address". In this case, PID and file name
to be looked up, perform role of "Logical address".
But, there is the big difference between
"Machida, Hiroyuki" <[EMAIL PROTECTED]> writes:
> Right, it looks like TLB, which holds cache "Physical addres"
> correponding to "Logical address". In this case, PID and file name
> to be looked up, perform role of "Logical address".
But, there is the big difference between hint table and TLB. T
On 8/31/05, Pekka Enberg <[EMAIL PROTECTED]> wrote:
> After finally understanding what you're doing, how about:
>
> static inline int hint_allocate(struct inode *dir)
> {
> loff_t *hints;
> int err = 0;
>
> if (!MSDOS_I(dir)->scan_hints)
Should read:
if (MSDOS_I(dir)->sc
On 8/31/05, Machida, Hiroyuki <[EMAIL PROTECTED]> wrote:
> How about this ?
>
> if (!MSDOS_I(dir)->scan_hints) {
> hints = kcalllo();
>
> down
> if (MSDOS_I(dir)->scan_hints) {
> up
> goto
On 8/31/05, Machida, Hiroyuki <[EMAIL PROTECTED]> wrote:
> > Please consider moving this check to callers. Conditional allocation
> > makes this bit strange API-wise. Or alternatively, give
> > hint_allocate() a better name.
>
> How about hint_allocate_conditional() ?
hint_get() sounds better to
Pekka Enberg wrote:
On 8/31/05, Machida, Hiroyuki <[EMAIL PROTECTED]> wrote:
+inline
+static int hint_index_body(const unsigned char *name, int name_len, int
check_null)
+{
+ int i;
+ int val = 0;
+ unsigned char *p = (unsigned char *) name;
+ int id = current->pid;
+
+
Hi,
Thank you checking code...
Pekka Enberg wrote:
Hi,
:
snip
This patch enables using hint information on scanning dir.
It achieves excellent performance with "ls -l" for over 1000 entries.
* fat-dirscan-with-hint_3.patch for linux 2.6.13
fs/fat/dir.c | 130
Hi,
Pekka Enberg wrote:
Hi,
On 8/31/05, Machida, Hiroyuki <[EMAIL PROTECTED]> wrote:
+inline
+static int hint_allocate(struct inode *dir)
+{
+ loff_t *hints;
+ int err = 0;
+
+ if (!MSDOS_I(dir)->scan_hints) {
+ hints = kcalloc(FAT_SCAN_NWAY, sizeof(loff_t), GF
Hi,
On 8/31/05, Machida, Hiroyuki <[EMAIL PROTECTED]> wrote:
>
> Sorry, I send out wrong version. I attached the latest patch to 2.6.13.
>
> OGAWA Hirofumi wrote:
> > "Machida, Hiroyuki" <[EMAIL PROTECTED]> writes:
> >
> >
> >>Here is a revised version of dirent scan patch, mentioned at
> >>fol
On 8/31/05, Machida, Hiroyuki <[EMAIL PROTECTED]> wrote:
> +inline
> +static int hint_index_body(const unsigned char *name, int name_len, int
> check_null)
> +{
> + int i;
> + int val = 0;
> + unsigned char *p = (unsigned char *) name;
> + int id = current->pid;
> +
> +
Hi,
On 8/31/05, Machida, Hiroyuki <[EMAIL PROTECTED]> wrote:
> +inline
> +static int hint_allocate(struct inode *dir)
> +{
> + loff_t *hints;
> + int err = 0;
> +
> + if (!MSDOS_I(dir)->scan_hints) {
> + hints = kcalloc(FAT_SCAN_NWAY, sizeof(loff_t), GFP_KERNEL);
>
Sorry, I send out wrong version. I attached the latest patch to 2.6.13.
OGAWA Hirofumi wrote:
"Machida, Hiroyuki" <[EMAIL PROTECTED]> writes:
Here is a revised version of dirent scan patch, mentioned at
following E-mail.
This patch addresses performance damages on "ls | xargs xxx" and
reve
14 matches
Mail list logo