mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Vishnu Reddy <busanna.reddy@oss.qualcomm.com>
To: Yijie Yang <yijie.yang@oss.qualcomm.com>,
	Vikash Garodia <vikash.garodia@oss.qualcomm.com>,
	"Joerg Roedel (AMD)" <joro@8bytes.org>,
	Vasant Hegde <vasant.hegde@amd.com>,
	Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>,
	Will Deacon <will@kernel.org>,
	Robin Murphy <robin.murphy@arm.com>,
	David Woodhouse <dwmw2@infradead.org>,
	Lu Baolu <baolu.lu@linux.intel.com>,
	Jean-Philippe Brucker <jpb@kernel.org>,
	Rob Herring <robh@kernel.org>,
	Saravana Kannan <saravanak@kernel.org>,
	Tomasz Nowicki <tomasz.nowicki@caviumnetworks.com>,
	Eric Auger <eric.auger@redhat.com>,
	Bryan O'Donoghue <bryan.odonoghue@linaro.org>,
	Joerg Roedel <jroedel@suse.de>,
	Srinath Mannam <srinath.mannam@broadcom.com>,
	Oza Pawandeep <quic_poza@quicinc.com>,
	Lorenzo Pieralisi <lpieralisi@kernel.org>,
	Bjorn Helgaas <bhelgaas@google.com>,
	Alex Williamson <alex@shazbot.org>
Cc: iommu@lists.linux.dev, linux-kernel@vger.kernel.org,
	virtualization@lists.linux.dev, devicetree@vger.kernel.org,
	kvm@vger.kernel.org, linux-arm-msm@vger.kernel.org,
	linux-media@vger.kernel.org,
	Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Subject: Re: [PATCH v7 5/5] iommu/of: Add support for "iommu-ranges" on a device node
Date: Tue, 17 Nov 2026 10:58:34 +0530	[thread overview]
Message-ID: <0a4efa52-f05a-cbf2-52da-6835343efc4d@oss.qualcomm.com> (raw)
In-Reply-To: <68bb5fe0-9005-4517-aec0-c39021a5b24f@oss.qualcomm.com>


On 11/13/2026 3:06 PM, Yijie Yang wrote:
>
>
> On 11/7/2026 1:14 AM, Vikash Garodia wrote:
>> Reserving an IOVA address using "iommu-addresses" requires it to be
>> placed in the reserved-memory node. But when "iommu-addresses" is the
>> only property being described and there is no backing "reg" (i.e. no
>> actual reserved system memory), it does not really belong under
>> /reserved-memory. A device IOVA range is specific to its own address
>> space, and does not describe the system physical memory to make it
>> qualify under reserved-memory. Given this, place "iommu-addresses"
>> inside reserved-memory only when it is paired with a "reg", otherwise,
>> define it within the device own node. It was discussed here:
>> https://lore.kernel.org/all/662f7093-fb0a-4564-9ca0-98e03e68ba9a@kernel.org
>>
>> Existing "iommu-addresses" property expects a phandle, which does
>> not make sense when the property is defined within the device node.
>> "iommu-ranges" was introduced to specify the device specific IOVA ranges
>> when there is no backing "reg".
>>
>> Suggested-by: Rob Herring <robh@kernel.org>
>> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
>> Co-developed-by: Vishnu Reddy <busanna.reddy@oss.qualcomm.com>
>> Signed-off-by: Vishnu Reddy <busanna.reddy@oss.qualcomm.com>
>> Signed-off-by: Vikash Garodia <vikash.garodia@oss.qualcomm.com>
>> ---
>>   drivers/iommu/of_iommu.c | 107 +++++++++++++++++++++++++++++++++++++++++++++++
>>   1 file changed, 107 insertions(+)
>>
>> diff --git a/drivers/iommu/of_iommu.c b/drivers/iommu/of_iommu.c
>> index
>> 00b02c19f27e195fc36c4bc1d91253255147c9ab..e133d3b346feecac7e2bfaec44e4882060ee89ec
>> 100644
>> --- a/drivers/iommu/of_iommu.c
>> +++ b/drivers/iommu/of_iommu.c
>> @@ -183,6 +183,111 @@ iommu_resv_region_get_type(struct device *dev,
>>       return IOMMU_RESV_RESERVED;
>>   }
>>   +/**
>> + * of_iommu_derive_resv_regions - derive reserved regions which
>> + *                                are outside of iommu-ranges
>> + * @dev: device for which to get reserved regions
>> + * @list: reserved region list
>> + *
>> + * A device can describe its own usable IOVA ranges directly on its node
>> + * via "iommu-ranges". Everything not under those ranges is derived
>> + * as a reserved region so the IOMMU allocator won't use it. Entries may
>> + * appear in any order in the property.
>> + */
>> +static void of_iommu_derive_resv_regions(struct device *dev, struct
>> list_head *list)
>> +{
>> +    struct of_iommu_range {
>> +        struct list_head node;
>> +        phys_addr_t start;
>> +        phys_addr_t end;
>> +    } *pos, *new, *next_range;
>> +    int size, prot = IOMMU_READ | IOMMU_WRITE;
>> +    struct iommu_resv_region *region;
>> +    const __be32 *maps, *end;
>> +    struct list_head *prev;
>> +    phys_addr_t next = 0;
>> +    LIST_HEAD(ranges);
>> +
>> +    maps = of_get_property(dev->of_node, "iommu-ranges", &size);
>> +    if (!maps)
>> +        return;
>> +
>> +    end = maps + size / sizeof(__be32);
>> +
>> +    while (maps < end) {
>> +        phys_addr_t iova, iova_end;
>> +        resource_size_t length;
>> +
>> +        maps = of_translate_dma_region(dev->of_node, maps, &iova, &length);
>> +        if (!maps) {
>> +            dev_err(dev, "%pOF: failed to parse iommu-ranges\n",
>> +                dev->of_node);
>> +            goto exit;
>> +        }
>> +
>> +        if (!length)
>> +            continue;
>> +
>> +        iova_end = iova + length - 1;
>> +
>> +        if (iova_end < iova) {
>> +            dev_err(dev, "%pOF: iommu-ranges overflows address space\n",
>> +                dev->of_node);
>> +            goto exit;
>> +        }
>> +
>> +        prev = &ranges;
>> +        list_for_each_entry(pos, &ranges, node) {
>> +            if (pos->start > iova)
>> +                break;
>> +            prev = &pos->node;
>> +        }
>> +
>> +        new = kmalloc_obj(*new);
>> +        if (!new)
>> +            goto exit;
>> +
>> +        new->start = iova;
>> +        new->end = iova_end;
>> +        list_add(&new->node, prev);
>> +    }
>> +
>> +    if (list_empty(&ranges))
>> +        return;
>> +
>> +    if (of_dma_is_coherent(dev->of_node))
>> +        prot |= IOMMU_CACHE;
>> +
>> +    list_for_each_entry_safe(pos, next_range, &ranges, node) {
>> +        if (pos->start > next) {
>> +            region = iommu_alloc_resv_region(next, pos->start - next, prot,
>> +                             IOMMU_RESV_RESERVED, GFP_KERNEL);
>> +            if (region)
>> +                list_add_tail(&region->list, list);
>
> Why keep looping here after allocation failure?

This matches the existing "iommu-addresses" handling. We continue after an
allocation failure since later allocations may still succeed, preserving as
many DT-specified restrictions as possible.

>
>> +        }
>> +
>> +        if (pos->end >= PHYS_ADDR_MAX - 1)
>> +            goto exit;
>> +
>> +        if (pos->end >= next)
>> +            next = pos->end + 1;
>> +
>> +        list_del(&pos->node);
>> +        kfree(pos);
>> +    }
>> +
>> +    region = iommu_alloc_resv_region(next, PHYS_ADDR_MAX - next,
>> +                     prot, IOMMU_RESV_RESERVED, GFP_KERNEL);
>> +    if (region)
>> +        list_add_tail(&region->list, list);
>> +
>> +exit:
>> +    list_for_each_entry_safe(pos, next_range, &ranges, node) {
>> +        list_del(&pos->node);
>> +        kfree(pos);
>> +    }
>> +}
>> +
>>   /**
>>    * of_iommu_get_resv_regions - reserved region driver helper for device tree
>>    * @dev: device for which to get reserved regions
>> @@ -262,5 +367,7 @@ void of_iommu_get_resv_regions(struct device *dev, struct
>> list_head *list)
>>               }
>>           }
>>       }
>> +
>> +    of_iommu_derive_resv_regions(dev, list);
>>   }
>>   EXPORT_SYMBOL(of_iommu_get_resv_regions);
>>
>

  reply	other threads:[~2026-11-17  5:28 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-06 17:14 [PATCH v7 0/5] of_iommu: add support for "iommu-ranges" Vikash Garodia
2026-11-06 17:14 ` [PATCH v7 1/5] iommu/dma: Keep reserved regions below the IOVA rbtree anchor Vikash Garodia
2026-11-11 13:03   ` Dmitry Baryshkov
2026-11-13 12:08   ` Krzysztof Kozlowski
2026-11-06 17:14 ` [PATCH v7 2/5] iommu: Use resource_size_t for reserved region length Vikash Garodia
2026-11-10 17:28   ` Rob Herring (Arm)
2026-11-11 13:02   ` Dmitry Baryshkov
2026-11-13 12:11   ` Krzysztof Kozlowski
2026-11-06 17:14 ` [PATCH v7 3/5] iommu/of: Make OF_IOMMU depend on OF_ADDRESS Vikash Garodia
2026-11-11 13:03   ` Dmitry Baryshkov
2026-11-12 13:40     ` Vikash Garodia
2026-11-13 12:37   ` Krzysztof Kozlowski
2026-11-06 17:14 ` [PATCH v7 4/5] iommu/of: Require reg for iommu-addresses Vikash Garodia
2026-11-11 12:02   ` Bryan O'Donoghue
2026-11-12 14:04     ` Vikash Garodia
2026-11-12 14:19       ` Bryan O'Donoghue
2026-11-13 13:11   ` Krzysztof Kozlowski
2026-11-06 17:14 ` [PATCH v7 5/5] iommu/of: Add support for "iommu-ranges" on a device node Vikash Garodia
2026-11-13  9:36   ` Yijie Yang
2026-11-17  5:28     ` Vishnu Reddy [this message]
2026-11-13 13:20   ` Krzysztof Kozlowski
2026-11-11 11:04 ` [PATCH v7 0/5] of_iommu: add support for "iommu-ranges" Vinayak Katoch
2026-11-11 13:01 ` Dmitry Baryshkov

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=0a4efa52-f05a-cbf2-52da-6835343efc4d@oss.qualcomm.com \
    --to=busanna.reddy@oss.qualcomm.com \
    --cc=alex@shazbot.org \
    --cc=baolu.lu@linux.intel.com \
    --cc=bhelgaas@google.com \
    --cc=bryan.odonoghue@linaro.org \
    --cc=devicetree@vger.kernel.org \
    --cc=dmitry.baryshkov@oss.qualcomm.com \
    --cc=dwmw2@infradead.org \
    --cc=eric.auger@redhat.com \
    --cc=iommu@lists.linux.dev \
    --cc=joro@8bytes.org \
    --cc=jpb@kernel.org \
    --cc=jroedel@suse.de \
    --cc=kvm@vger.kernel.org \
    --cc=linux-arm-msm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-media@vger.kernel.org \
    --cc=lpieralisi@kernel.org \
    --cc=quic_poza@quicinc.com \
    --cc=robh@kernel.org \
    --cc=robin.murphy@arm.com \
    --cc=saravanak@kernel.org \
    --cc=srinath.mannam@broadcom.com \
    --cc=suravee.suthikulpanit@amd.com \
    --cc=tomasz.nowicki@caviumnetworks.com \
    --cc=vasant.hegde@amd.com \
    --cc=vikash.garodia@oss.qualcomm.com \
    --cc=virtualization@lists.linux.dev \
    --cc=will@kernel.org \
    --cc=yijie.yang@oss.qualcomm.com \
    /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

all inboxes | Powered by JetHome®