LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: "Bryan O'Donoghue" <pure.logic@nexus-software.ie>
To: "Ong, Boon Leong" <boon.leong.ong@intel.com>
Cc: "tglx@linutronix.de" <tglx@linutronix.de>,
	"mingo@redhat.com" <mingo@redhat.com>,
	"hpa@zytor.com" <hpa@zytor.com>,
	"x86@kernel.org" <x86@kernel.org>,
	"dvhart@infradead.org" <dvhart@infradead.org>,
	"andy.shevchenko@gmail.com" <andy.shevchenko@gmail.com>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v2 1/1] x86: Add Isolated Memory Regions for Quark X1000
Date: Sat, 24 Jan 2015 19:52:07 +0000	[thread overview]
Message-ID: <54C3F7E7.6090306@nexus-software.ie> (raw)
In-Reply-To: <AF233D1473C1364ABD51D28909A1B1B73270D4C5@PGSMSX105.gar.corp.intel.com>

On 24/01/15 01:48, Ong, Boon Leong wrote:

Skipping stuff I agree with.

>  From V1 comment:
> Suggest to add a statement on 3 different types of IMR: General IMR, Host Memory
> I/O Boundary IMR & System Management Mode IMR. Then, call out that this patch
> is meant to support general IMR type only.

Hmm - There's no mention of grouping like that in the documentation, nor 
in released silicon - to my knowledge.

Also why do you want a statement added saying that it supports CPU only 
mode ?

This patch will support adding IMRs for SMM mode - if calling code wants 
do do that - it's just imr_add_range(base, size, SMM, SMM);

Same thing with virtual-channels, RMU, etc.


>> +	ret = imr_check_range(base, size);
>> +	if (ret)
>> +		return ret;
>> +
>> +	if (size < IMR_ALIGN)
>> +		return -EINVAL;
> I believe this is redundant because imr_check_range() has test for (size & IMR_MASK)
> which means if the size is indeed smaller than 0x400, the test will caught it anyway.

Nope.

(0 & 0x3FF) == 0

We need to bounds check for a zero size.

I'll change it to

if (size == 0)
	return -EINVAL;

to avoid confusion.

>> +
>> +	/* Tweak the size value */
>> +	size = imr_fixup_size(size);
>> +	pr_debug("IMR %d phys 0x%08lx-0x%08lx rmask 0x%08x wmask
>> 0x%08x\n",
>> +		reg, base, end, rmask, wmask);
> Do we want to account for the 'size fixup' above on 'end'
>> +
>> +	/* Allocate IMR */
>> +	imr.addr_lo = phys_to_imr(base);
>> +	imr.addr_hi = phys_to_imr(end);
>
> The fix-up size above is never factored here ...
> 'end-size' should be the correct one

hmmm.

The correct fix is

size = imr_fixup_size(size);
end = base + size;

>> +	} else {
>> +		/* Search for match based on address range */
>> +		for (i = 0; i < imr_dev.max_imr; i++) {
>> +			ret = imr_read(reg, &imr);
> A serious bug here.... 'reg' should be 'i' . We enter this branch if reg=-1
> Is there a miss in your test case?

Hmm you're right.

Turns out there's only the one test case for imr_del_range();

Good catch.

--
BOD


      parent reply	other threads:[~2015-01-24 19:52 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-01-21 18:46 [PATCH v2 0/1] x86: Add IMR support to Quark/Galileo Bryan O'Donoghue
2015-01-21 18:46 ` [PATCH v2 1/1] x86: Add Isolated Memory Regions for Quark X1000 Bryan O'Donoghue
2015-01-21 20:57   ` Andy Shevchenko
2015-01-22  1:27     ` Bryan O'Donoghue
2015-01-22  8:59       ` Andy Shevchenko
2015-01-22  9:43         ` Bryan O'Donoghue
2015-01-22 11:24   ` Thomas Gleixner
2015-01-22 11:38     ` Bryan O'Donoghue
2015-01-22 15:02       ` Bryan O'Donoghue
2015-01-22 15:15         ` Bryan O'Donoghue
2015-01-22 16:28           ` Darren Hart
2015-01-22 19:50           ` Thomas Gleixner
2015-01-24  1:48   ` Ong, Boon Leong
2015-01-24 11:02     ` Andy Shevchenko
2015-01-24 21:56       ` Bryan O'Donoghue
2015-01-24 21:58         ` Bryan O'Donoghue
2015-01-24 19:52     ` Bryan O'Donoghue [this message]

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=54C3F7E7.6090306@nexus-software.ie \
    --to=pure.logic@nexus-software.ie \
    --cc=andy.shevchenko@gmail.com \
    --cc=boon.leong.ong@intel.com \
    --cc=dvhart@infradead.org \
    --cc=hpa@zytor.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=tglx@linutronix.de \
    --cc=x86@kernel.org \
    --subject='Re: [PATCH v2 1/1] x86: Add Isolated Memory Regions for Quark X1000' \
    /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

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).