| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <7186caa3-13c0-4629-a07f-162c00138ad9@linux.intel.com>
Date: Mon, 15 Dec 2025 11:30:39 +0800
From: Baolu Lu <baolu.lu@...ux.intel.com>
To: Mostafa Saleh <smostafa@...gle.com>
Cc: linux-mm@...ck.org, iommu@...ts.linux.dev, linux-kernel@...r.kernel.org,
linux-doc@...r.kernel.org, corbet@....net, joro@...tes.org, will@...nel.org,
robin.murphy@....com, akpm@...ux-foundation.org, vbabka@...e.cz,
surenb@...gle.com, mhocko@...e.com, jackmanb@...gle.com, hannes@...xchg.org,
ziy@...dia.com, david@...hat.com, lorenzo.stoakes@...cle.com,
Liam.Howlett@...cle.com, rppt@...nel.org, xiaqinxin@...wei.com,
rdunlap@...radead.org
Subject: Re: [PATCH v4 2/4] iommu: Add calls for IOMMU_DEBUG_PAGEALLOC
On 12/13/25 02:44, Mostafa Saleh wrote:
> On Fri, Dec 12, 2025 at 10:33:20AM +0800, Baolu Lu wrote:
>> On 12/11/25 20:59, Mostafa Saleh wrote:
>>> Add calls for the new iommu debug config IOMMU_DEBUG_PAGEALLOC:
>>> - iommu_debug_init: Enable the debug mode if configured by the user.
>>> - iommu_debug_map: Track iommu pages mapped, using physical address.
>>> - iommu_debug_unmap_begin: Track start of iommu unmap operation, with
>>> IOVA and size.
>>> - iommu_debug_unmap_end: Track the end of unmap operation, passing the
>>> actual unmapped size versus the tracked one at unmap_begin.
>>>
>>> We have to do the unmap_begin/end as once pages are unmapped we lose
>>> the information of the physical address.
>>> This is racy, but the API is racy by construction as it uses refcounts
>>> and doesn't attempt to lock/synchronize with the IOMMU API as that will
>>> be costly, meaning that possibility of false negative exists.
>>>
>>> Signed-off-by: Mostafa Saleh<smostafa@...gle.com>
>>> ---
>>> drivers/iommu/iommu-debug-pagealloc.c | 28 +++++++++++++
>>> drivers/iommu/iommu-priv.h | 58 +++++++++++++++++++++++++++
>>> drivers/iommu/iommu.c | 11 ++++-
>>> include/linux/iommu-debug-pagealloc.h | 1 +
>>> 4 files changed, 96 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/drivers/iommu/iommu-debug-pagealloc.c b/drivers/iommu/iommu-debug-pagealloc.c
>>> index 4022e9af7f27..1d343421da98 100644
>>> --- a/drivers/iommu/iommu-debug-pagealloc.c
>>> +++ b/drivers/iommu/iommu-debug-pagealloc.c
>>> @@ -5,11 +5,15 @@
>>> * IOMMU API debug page alloc sanitizer
>>> */
>>> #include <linux/atomic.h>
>>> +#include <linux/iommu.h>
>>> #include <linux/iommu-debug-pagealloc.h>
>>> #include <linux/kernel.h>
>>> #include <linux/page_ext.h>
>>> +#include "iommu-priv.h"
>>> +
>>> static bool needed;
>>> +DEFINE_STATIC_KEY_FALSE(iommu_debug_initialized);
>>> struct iommu_debug_metadata {
>>> atomic_t ref;
>>> @@ -25,6 +29,30 @@ struct page_ext_operations page_iommu_debug_ops = {
>>> .need = need_iommu_debug,
>>> };
>>> +void __iommu_debug_map(struct iommu_domain *domain, phys_addr_t phys, size_t size)
>>> +{
>>> +}
>>> +
>>> +void __iommu_debug_unmap_begin(struct iommu_domain *domain,
>>> + unsigned long iova, size_t size)
>>> +{
>>> +}
>>> +
>>> +void __iommu_debug_unmap_end(struct iommu_domain *domain,
>>> + unsigned long iova, size_t size,
>>> + size_t unmapped)
>>> +{
>>> +}
>>> +
>>> +void iommu_debug_init(void)
>>> +{
>>> + if (!needed)
>>> + return;
>>> +
>>> + pr_info("iommu: Debugging page allocations, expect overhead or disable iommu.debug_pagealloc");
>>> + static_branch_enable(&iommu_debug_initialized);
>>> +}
>>> +
>>> static int __init iommu_debug_pagealloc(char *str)
>>> {
>>> return kstrtobool(str, &needed);
>>> diff --git a/drivers/iommu/iommu-priv.h b/drivers/iommu/iommu-priv.h
>>> index c95394cd03a7..aaffad5854fc 100644
>>> --- a/drivers/iommu/iommu-priv.h
>>> +++ b/drivers/iommu/iommu-priv.h
>>> @@ -5,6 +5,7 @@
>>> #define __LINUX_IOMMU_PRIV_H
>>> #include <linux/iommu.h>
>>> +#include <linux/iommu-debug-pagealloc.h>
>>> #include <linux/msi.h>
>>> static inline const struct iommu_ops *dev_iommu_ops(struct device *dev)
>>> @@ -65,4 +66,61 @@ static inline int iommufd_sw_msi(struct iommu_domain *domain,
>>> int iommu_replace_device_pasid(struct iommu_domain *domain,
>>> struct device *dev, ioasid_t pasid,
>>> struct iommu_attach_handle *handle);
>>> +
>>> +#ifdef CONFIG_IOMMU_DEBUG_PAGEALLOC
>>> +
>>> +void __iommu_debug_map(struct iommu_domain *domain, phys_addr_t phys,
>>> + size_t size);
>>> +void __iommu_debug_unmap_begin(struct iommu_domain *domain,
>>> + unsigned long iova, size_t size);
>>> +void __iommu_debug_unmap_end(struct iommu_domain *domain,
>>> + unsigned long iova, size_t size, size_t unmapped);
>>> +
>>> +static inline void iommu_debug_map(struct iommu_domain *domain,
>>> + phys_addr_t phys, size_t size)
>>> +{
>>> + if (static_branch_unlikely(&iommu_debug_initialized))
>>> + __iommu_debug_map(domain, phys, size);
>>> +}
>>> +
>>> +static inline void iommu_debug_unmap_begin(struct iommu_domain *domain,
>>> + unsigned long iova, size_t size)
>>> +{
>>> + if (static_branch_unlikely(&iommu_debug_initialized))
>>> + __iommu_debug_unmap_begin(domain, iova, size);
>>> +}
>>> +
>>> +static inline void iommu_debug_unmap_end(struct iommu_domain *domain,
>>> + unsigned long iova, size_t size,
>>> + size_t unmapped)
>>> +{
>>> + if (static_branch_unlikely(&iommu_debug_initialized))
>>> + __iommu_debug_unmap_end(domain, iova, size, unmapped);
>>> +}
>> I am wondering whether it would be better if we move iommu_debug_map()
>> to iommu-debug-pagealloc.c,
>>
>> void iommu_debug_map(struct iommu_domain *domain,
>> phys_addr_t phys, size_t size)
>> {
>> if (static_branch_likely(&iommu_debug_initialized))
>> __iommu_debug_map(domain, phys, size);
>> }
>>
>> (Does it make sense to use static_branch_likely() here? Normally, people
>> who enable CONFIG_IOMMU_DEBUG_PAGEALLOC would want to use this
>> debugging feature. Or not?)
>>
>> So that ...
> This actually was the v1 implementation [1], but Jörg suggested to move
> it to a header file as a function call would have an overhead if this
> feautre is disabled.
>
> I believe the priority would be to keep the performance overhead minimal
> with CONFIG_IOMMU_DEBUG_PAGEALLOC and the commandline disabled, so people
> can run with the config in production and only enable the commandline
> it to debug problems, without having overhead on the typical case.
Okay, fair enough.
Thanks,
baolu
Powered by blists - more mailing lists