From: Joerg Roedel <jroedel@suse.de>
To: Qian Cai <cai@lca.pw>
Cc: don.brace@microsemi.com, iommu@lists.linux-foundation.org,
esc.storagedev@microsemi.com, hch@lst.de,
linux-kernel@vger.kernel.org
Subject: Re: [RFC PATCH] iommu/amd: fix a race in increase_address_space()
Date: Thu, 5 Sep 2019 13:43:39 +0200 [thread overview]
Message-ID: <20190905114339.GC5457@suse.de> (raw)
In-Reply-To: <1567632262-21284-1-git-send-email-cai@lca.pw>
Hi Qian,
On Wed, Sep 04, 2019 at 05:24:22PM -0400, Qian Cai wrote:
> if (domain->mode == PAGE_MODE_6_LEVEL)
> /* address space already 64 bit large */
> return false;
>
> This gives a clue that there must be a race between multiple concurrent
> threads in increase_address_space().
Thanks for tracking this down, there is a race indeed.
> + mutex_lock(&domain->api_lock);
> *dma_addr = __map_single(dev, dma_dom, page_to_phys(page),
> size, DMA_BIDIRECTIONAL, dma_mask);
> + mutex_unlock(&domain->api_lock);
>
> if (*dma_addr == DMA_MAPPING_ERROR)
> goto out_free;
> @@ -2696,7 +2698,9 @@ static void free_coherent(struct device *dev, size_t size,
>
> dma_dom = to_dma_ops_domain(domain);
>
> + mutex_lock(&domain->api_lock);
> __unmap_single(dma_dom, dma_addr, size, DMA_BIDIRECTIONAL);
> + mutex_unlock(&domain->api_lock);
But I think the right fix is to lock the operation in
increase_address_space() directly, and not the calls around it, like in
the diff below. It is untested, so can you please try it and report back
if it fixes your issue?
diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c
index b607a92791d3..1ff705f16239 100644
--- a/drivers/iommu/amd_iommu.c
+++ b/drivers/iommu/amd_iommu.c
@@ -1424,18 +1424,21 @@ static void free_pagetable(struct protection_domain *domain)
* another level increases the size of the address space by 9 bits to a size up
* to 64 bits.
*/
-static bool increase_address_space(struct protection_domain *domain,
+static void increase_address_space(struct protection_domain *domain,
gfp_t gfp)
{
+ unsigned long flags;
u64 *pte;
- if (domain->mode == PAGE_MODE_6_LEVEL)
+ spin_lock_irqsave(&domain->lock, flags);
+
+ if (WARN_ON_ONCE(domain->mode == PAGE_MODE_6_LEVEL))
/* address space already 64 bit large */
- return false;
+ goto out;
pte = (void *)get_zeroed_page(gfp);
if (!pte)
- return false;
+ goto out;
*pte = PM_LEVEL_PDE(domain->mode,
iommu_virt_to_phys(domain->pt_root));
@@ -1443,7 +1446,10 @@ static bool increase_address_space(struct protection_domain *domain,
domain->mode += 1;
domain->updated = true;
- return true;
+out:
+ spin_unlock_irqrestore(&domain->lock, flags);
+
+ return;
}
static u64 *alloc_pte(struct protection_domain *domain,
_______________________________________________
iommu mailing list
iommu@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/iommu
next prev parent reply other threads:[~2019-09-05 11:43 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-09-04 21:24 [RFC PATCH] iommu/amd: fix a race in increase_address_space() Qian Cai
2019-09-05 11:43 ` Joerg Roedel [this message]
2019-09-05 20:57 ` Qian Cai
2019-09-06 8:56 ` [PATCH] iommu/amd: Fix " Joerg Roedel
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=20190905114339.GC5457@suse.de \
--to=jroedel@suse.de \
--cc=cai@lca.pw \
--cc=don.brace@microsemi.com \
--cc=esc.storagedev@microsemi.com \
--cc=hch@lst.de \
--cc=iommu@lists.linux-foundation.org \
--cc=linux-kernel@vger.kernel.org \
/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
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).