Kernel KVM virtualization development
 help / color / mirror / Atom feed
From: fengchengwen <fengchengwen@huawei.com>
To: Alex Williamson <alex@shazbot.org>
Cc: <jgg@ziepe.ca>, <helgaas@kernel.org>,
	<wathsala.vithanage@arm.com>, <wei.huang2@amd.com>,
	<zhipingz@meta.com>, <wangzhou1@hisilicon.com>,
	<wangyushan12@huawei.com>, <liuyonglong@huawei.com>,
	<kvm@vger.kernel.org>, <linux-pci@vger.kernel.org>
Subject: Re: [PATCH v19 13/18] vfio/pci: Introduce VFIO_DEVICE_FEATURE_TPH family uapi for PCI TPH control
Date: Thu, 9 Jul 2026 19:56:21 +0800	[thread overview]
Message-ID: <548acc4a-c9c2-458e-8f0a-0822f7a7c2ab@huawei.com> (raw)
In-Reply-To: <20260708173101.236cd0b9@shazbot.org>

Hi Alex,

On 7/9/2026 7:31 AM, Alex Williamson wrote:
> On Thu, 2 Jul 2026 20:42:19 +0800
> Chengwen Feng <fengchengwen@huawei.com> wrote:
> 
>> Add user API definitions for PCI TPH device feature set, including three
>> feature IDs:
>> - VFIO_DEVICE_FEATURE_TPH: Global TPH enable and capability query
>> - VFIO_DEVICE_FEATURE_TPH_RESOLVE: Resolve PH/ST from specified source
>>   (GET only)
>> - VFIO_DEVICE_FEATURE_TPH_ST: Batch program ST table entries (SET only)
>>
>> Signed-off-by: Chengwen Feng <fengchengwen@huawei.com>
>> ---
>>  include/uapi/linux/vfio.h | 62 +++++++++++++++++++++++++++++++++++++++
>>  1 file changed, 62 insertions(+)
>>
>> diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h
>> index 5de618a3a5ee..ed04b6280cd6 100644
>> --- a/include/uapi/linux/vfio.h
>> +++ b/include/uapi/linux/vfio.h
>> @@ -1534,6 +1534,68 @@ struct vfio_device_feature_dma_buf {
>>   */
>>  #define VFIO_DEVICE_FEATURE_MIG_PRECOPY_INFOv2  12
>>  
>> +/* PCIe TPH device feature definitions for VFIO_DEVICE_FEATURE ioctl */
>> +#define VFIO_DEVICE_FEATURE_TPH		13
>> +#define VFIO_DEVICE_FEATURE_TPH_RESOLVE	14
>> +#define VFIO_DEVICE_FEATURE_TPH_ST	15
>> +
>> +/*
>> + * VFIO_DEVICE_FEATURE_TPH - Control and query PCI TPH capabilities
>> + *
>> + * SET: Opt-in to TPH feature; flags must be zero.
> 
> We clearly need to define what it actually means to opt-in, as this
> series is not implementing it as it should be defined.  The user needs
> to opt-in via this feature before we modify the device ABI in any way.
> The extent to which the user may enable TPH features is bound both by
> the device capabilities and the module parameters set by the system
> admin.

I will add full documentation describing the mandatory opt-in rule: users
must call SET on this feature before any TPH virtualization or TPH_RESOLVE/
TPH_ST ioctl access is allowed. Without opt-in, TPH config space registers
remain raw read-only and related TPH feature ioctls return -ENOTSUPP.
Also clarify the dual capability bound: exposed TPH functions are limited by
both hardware native TPH capability and host-wide tph module policy.

> 
>> + * GET: Return supported TPH capability bits in flags.
>> + */
>> +struct vfio_device_feature_tph {
>> +	__u32 flags;
>> +};
>> +
>> +/* Capability bits returned by GET on VFIO_DEVICE_FEATURE_TPH */
>> +#define VFIO_DEVICE_TPH_CAP_DMABUF	(1u << 0) /* DMABUF sources resolvable */
>> +#define VFIO_DEVICE_TPH_CAP_CPU		(1u << 1) /* CPU sources resolvable via _DSM */
>> +#define VFIO_DEVICE_TPH_CAP_LITERAL	(1u << 2) /* LITERAL ST source supported */
>> +
>> +/*
>> + * VFIO_DEVICE_FEATURE_TPH_RESOLVE - Resolve PH/ST tag value from source
>> + * Note: GET only
>> + */
>> +struct vfio_device_feature_tph_resolve {
>> +	__u32 flags;	/* IN: source type + namespace modifiers */
>> +	__u32 src;	/* IN: dma-buf fd or CPU ID based on flags */
>> +	__u8 valid;	/* OUT: bitmap indicating valid output fields */
>> +#define VFIO_DEVICE_TPH_VALID_PH (1u << 0) /* ph holds valid processing hint */
>> +#define VFIO_DEVICE_TPH_VALID_ST (1u << 1) /* st holds valid steering tag */
>> +	__u8 ph;	/* OUT: TPH processing hint */
>> +	__u16 st;	/* OUT: raw steering tag value */
>> +};
>> +
>> +/*
>> + * VFIO_DEVICE_FEATURE_TPH_ST - Batch program architected ST table entries
>> + * Return value: number of successfully written entries
>> + * Note: SET only
>> + */
>> +struct vfio_device_feature_tph_st {
>> +	__u32 flags;		/* IN: source type + namespace modifiers */
>> +	__u16 start;		/* IN: first ST table index / MSI-X vector */
>> +	__u16 count;		/* IN: number of contiguous entries to program */
>> +	__aligned_u64 dests;	/* IN: user pointer to array of source u32 values */
>> +};
>> +
>> +/* Common flags shared by TPH_RESOLVE and TPH_ST features */
>> +#define VFIO_DEVICE_TPH_SRC_NONE           (1u << 0) /* TPH_ST only: set ST=0 */
>> +#define VFIO_DEVICE_TPH_SRC_DMABUF	   (1u << 1) /* src = dma-buf fd */
>> +#define VFIO_DEVICE_TPH_SRC_CPU_VOLATILE   (1u << 2) /* src = volatile memory CPU ID */
>> +#define VFIO_DEVICE_TPH_SRC_CPU_PERSISTENT (1u << 3) /* src = persistent memory CPU ID */
>> +#define VFIO_DEVICE_TPH_SRC_LITERAL        (1u << 4) /* TPH_ST only: src = raw literal ST */
>> +#define VFIO_DEVICE_TPH_EXTENDED           (1u << 30) /* Use extended 16-bit ST namespace */
>> +#define VFIO_DEVICE_TPH_REQUIRE_ST         (1u << 31) /* TPH_ST only: fail if resolved ST=0 */
> 
> I don't see the value in jumping to the end of the flags field for
> these last two.  Adding DMABUF is a good change from the proposal, but
> I wouldn't go out of our way to try to preserve some grouping of source
> flags from modifiers.

I will move VFIO_DEVICE_TPH_EXTENDED and VFIO_DEVICE_TPH_REQUIRE_ST down
from bits 30/31 to lower unused positions, remove artificial separation
between source-type flags and modifier flags for cleaner grouping.
> 
> We also seem to have avoided any statement or decision to what extent
> our uAPI is defined by the current implementation.  Particularly the
> ordering of TPH_ST versus manipulating the control bits in config space
> are entirely undefined here.
> 
> We're also not addressing that the TPH_ST policy counters the overall
> feature policy to return 0/-errno and not updating the documentation to
> bring that to the forefront.  Thanks,

Two undefined behaviors will be explicitly documented:
- Execution order and priority between VFIO_DEVICE_FEATURE_TPH_ST batch
  programming and direct userspace writes to TPH_CTRL config registers;
- Error return rules when requested operation exceeds current tph_policy
  level (Literal mode blocked under policy <= 2 returns -ENOTSUPP).

Both will be fix in v20.

Thanks

> 
> Alex
> 
>> +
>> +/* Mask for mutually exclusive source type bits, only one may be set */
>> +#define VFIO_DEVICE_TPH_SRC_MASK	(VFIO_DEVICE_TPH_SRC_NONE | \
>> +					 VFIO_DEVICE_TPH_SRC_DMABUF | \
>> +					 VFIO_DEVICE_TPH_SRC_CPU_VOLATILE | \
>> +					 VFIO_DEVICE_TPH_SRC_CPU_PERSISTENT | \
>> +					 VFIO_DEVICE_TPH_SRC_LITERAL)
>> +
>>  /* -------- API for Type1 VFIO IOMMU -------- */
>>  
>>  /**
> 


  reply	other threads:[~2026-07-09 11:56 UTC|newest]

Thread overview: 71+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-07-02 12:42 [PATCH v19 00/18] vfio/pci: Add PCIe TPH support Chengwen Feng
2026-07-02 12:42 ` [PATCH v19 01/18] PCI/TPH: Fix pcie_tph_get_st_table_loc() field extraction Chengwen Feng
2026-07-02 12:51   ` sashiko-bot
2026-07-02 12:42 ` [PATCH v19 02/18] PCI/TPH: Fix tph_enabled concurrent update race by bitfield packing Chengwen Feng
2026-07-02 12:51   ` sashiko-bot
2026-07-02 12:42 ` [PATCH v19 03/18] PCI/TPH: Cache TPH requester capability at probe time Chengwen Feng
2026-07-02 12:55   ` sashiko-bot
2026-07-02 12:42 ` [PATCH v19 04/18] PCI/TPH: Refactor pcie_enable_tph & add explicit requester variant Chengwen Feng
2026-07-02 12:50   ` sashiko-bot
2026-07-08 23:33   ` Alex Williamson
2026-07-09  9:22     ` fengchengwen
2026-07-02 12:42 ` [PATCH v19 05/18] PCI/TPH: Refactor pcie_tph_get_cpu_st & add explicit variant Chengwen Feng
2026-07-02 12:56   ` sashiko-bot
2026-07-02 12:42 ` [PATCH v19 06/18] PCI/TPH: Expose the enabled TPH requester type Chengwen Feng
2026-07-02 12:49   ` sashiko-bot
2026-07-08 23:33   ` Alex Williamson
2026-07-09  9:30     ` fengchengwen
2026-07-02 12:42 ` [PATCH v19 07/18] PCI/TPH: Add pcie_tph_supported() helper to check TPH capability attributes Chengwen Feng
2026-07-02 12:53   ` sashiko-bot
2026-07-03  0:39     ` fengchengwen
2026-07-08 23:32   ` Alex Williamson
2026-07-09  9:36     ` fengchengwen
2026-07-02 12:42 ` [PATCH v19 08/18] PCI/TPH: Add pci_tph_dsm_supported() helper to detect device TPH ST _DSM Chengwen Feng
2026-07-02 12:55   ` sashiko-bot
2026-07-02 12:42 ` [PATCH v19 09/18] vfio/pci: Hide TPH capability when TPH is unsupported Chengwen Feng
2026-07-02 13:00   ` sashiko-bot
2026-07-03  0:36     ` fengchengwen
2026-07-08 23:32   ` Alex Williamson
2026-07-09 11:00     ` fengchengwen
2026-07-02 12:42 ` [PATCH v19 10/18] vfio/pci: Introduce tph policy parameter for staged TPH feature enablement Chengwen Feng
2026-07-02 12:50   ` sashiko-bot
2026-07-08 23:31   ` Alex Williamson
2026-07-09 11:29     ` fengchengwen
2026-07-02 12:42 ` [PATCH v19 11/18] vfio/pci: Virtualize PCIe TPH capability registers Chengwen Feng
2026-07-02 13:04   ` sashiko-bot
2026-07-03  0:51     ` fengchengwen
2026-07-08 23:31   ` Alex Williamson
2026-07-09 11:40     ` fengchengwen
2026-07-02 12:42 ` [PATCH v19 12/18] vfio/pci: Add dmabuf TPH metadata storage and fd query helper Chengwen Feng
2026-07-02 12:56   ` sashiko-bot
2026-07-03  0:53     ` fengchengwen
2026-07-08 23:31   ` Alex Williamson
2026-07-09 11:49     ` fengchengwen
2026-07-09 18:50       ` Alex Williamson
2026-07-02 12:42 ` [PATCH v19 13/18] vfio/pci: Introduce VFIO_DEVICE_FEATURE_TPH family uapi for PCI TPH control Chengwen Feng
2026-07-02 13:01   ` sashiko-bot
2026-07-03  0:57     ` fengchengwen
2026-07-08 23:31   ` Alex Williamson
2026-07-09 11:56     ` fengchengwen [this message]
2026-07-02 12:42 ` [PATCH v19 14/18] vfio/pci: Implement VFIO_DEVICE_FEATURE_TPH and valid TPH config write support Chengwen Feng
2026-07-02 13:04   ` sashiko-bot
2026-07-03  1:16     ` fengchengwen
2026-07-08 23:30   ` Alex Williamson
2026-07-09 12:16     ` fengchengwen
2026-07-02 12:42 ` [PATCH v19 15/18] vfio/pci: Implement TPH_RESOLVE feature for DMABUF and CPU source resolving Chengwen Feng
2026-07-02 13:00   ` sashiko-bot
2026-07-03  1:26     ` fengchengwen
2026-07-08 23:31   ` Alex Williamson
2026-07-09 12:20     ` fengchengwen
2026-07-02 12:42 ` [PATCH v19 16/18] vfio/pci: Implement TPH_ST feature for batch ST table programming Chengwen Feng
2026-07-02 13:04   ` sashiko-bot
2026-07-03  1:42     ` fengchengwen
2026-07-08 23:29   ` Alex Williamson
2026-07-09 12:36     ` fengchengwen
2026-07-09 13:49       ` Alex Williamson
2026-07-02 12:42 ` [PATCH v19 17/18] vfio/pci: Reset hardware TPH state on device enable/disable Chengwen Feng
2026-07-02 13:00   ` sashiko-bot
2026-07-08 23:29   ` Alex Williamson
2026-07-09 12:39     ` fengchengwen
2026-07-02 12:42 ` [PATCH v19 18/18] vfio/pci: Expose tph_policy via debugfs Chengwen Feng
2026-07-02 12:59   ` sashiko-bot

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=548acc4a-c9c2-458e-8f0a-0822f7a7c2ab@huawei.com \
    --to=fengchengwen@huawei.com \
    --cc=alex@shazbot.org \
    --cc=helgaas@kernel.org \
    --cc=jgg@ziepe.ca \
    --cc=kvm@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=liuyonglong@huawei.com \
    --cc=wangyushan12@huawei.com \
    --cc=wangzhou1@hisilicon.com \
    --cc=wathsala.vithanage@arm.com \
    --cc=wei.huang2@amd.com \
    --cc=zhipingz@meta.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