Re: [PATCH v5 03/18] ACPI / table: Count matched and successfully parsed entries without specifying max entries

From: Rafael J. Wysocki
Date: Mon Nov 24 2014 - 09:55:50 EST


On Monday, November 24, 2014 09:34:24 AM Tomasz Nowicki wrote:
> On 24.11.2014 02:45, Rafael J. Wysocki wrote:
> > On Friday, October 17, 2014 09:36:59 PM Hanjun Guo wrote:
> >> From: Tomasz Nowicki <tomasz.nowicki@xxxxxxxxxx>
> >>
> >> It is very useful to traverse all available table entries without max
> >> number of expected entries type. Current acpi_parse_entries()
> >> implementation gives that feature but it does not count those entries,
> >> it returns 0 instead, so fix it to count matched and successfully
> >> entries and return it.
> >
> > Hmm. I guess that the goal is for count to only be incremented when the
> > condition is satisfied entirely, while without the patch it may be incremented
> > even if that isn't the case.
>
> That would be our goal if patch would look like:
> - && (!max_entries || count++ < max_entries))
> + && (!max_entries && count++ < max_entries)) {
> but then we can not walk through all available entries (with max_entries==0)

No, that's not what I was trying to say. :-)

> >
> > I'm not sure how that is related to the above paragraph, however.
> >
>
> Previous changelog is not clear, let me rewrite it:
>
> acpi_parse_entries() allows to traverse all available table entries (aka
> subtables) by passing max_entries parameter equal to 0. But for that use
> case acpi_parse_entries() does not inform caller how many entries were
> matched and for how many entries handler was run against. That patch is
> going to fix it.

Do I understand correctly that count is only ever incremented by current code
if max_entries is different from 0?


> >> NOTE: This change has no impact to x86 and ia64 archs since existing code
> >> checks for error occurrence only (acpi_parse_entries(...,0) < 0).
> >>
> >> Acked-by: Grant Likely <grant.likely@xxxxxxxxxx>
> >> Signed-off-by: Tomasz Nowicki <tomasz.nowicki@xxxxxxxxxx>
> >> Signed-off-by: Hanjun Guo <hanjun.guo@xxxxxxxxxx>
> >> ---
> >> drivers/acpi/tables.c | 5 ++++-
> >> 1 file changed, 4 insertions(+), 1 deletion(-)
> >>
> >> diff --git a/drivers/acpi/tables.c b/drivers/acpi/tables.c
> >> index 21ae521..b18e45e 100644
> >> --- a/drivers/acpi/tables.c
> >> +++ b/drivers/acpi/tables.c
> >> @@ -225,10 +225,13 @@ acpi_parse_entries(unsigned long table_size,
> >> while (((unsigned long)entry) + sizeof(struct acpi_subtable_header) <
> >> table_end) {
> >> if (entry->type == entry_id
> >> - && (!max_entries || count++ < max_entries))
> >> + && (!max_entries || count < max_entries)) {
> >> if (handler(entry, table_end))
> >> return -EINVAL;
> >>
> >> + count++;
> >> + }
> >> +
> >> /*
> >> * If entry->length is 0, break from this loop to avoid
> >> * infinite loop.
> >>
> >
> --
> To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html

--
I speak only for myself.
Rafael J. Wysocki, Intel Open Source Technology Center.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/