From: Lu Baolu <baolu.lu@linux.intel.com> To: "Tian, Kevin" <kevin.tian@intel.com>, Joerg Roedel <joro@8bytes.org>, Steve Wahl <steve.wahl@hpe.com> Cc: baolu.lu@linux.intel.com, David Woodhouse <dwmw2@infradead.org>, Jerry Snitselaar <jsnitsel@redhat.com>, Mike Travis <mike.travis@hpe.com>, Dimitri Sivanich <sivanich@hpe.com>, "Anderson, Russ" <russ.anderson@hpe.com>, "iommu@lists.linux.dev" <iommu@lists.linux.dev>, "iommu@lists.linux-foundation.org" <iommu@lists.linux-foundation.org>, "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org> Subject: Re: [PATCH v1 3/6] iommu/vt-d: Refactor iommu information of each domain Date: Fri, 1 Jul 2022 10:35:50 +0800 [thread overview] Message-ID: <595c314b-2d16-aeef-d808-d1cb713ba175@linux.intel.com> (raw) In-Reply-To: <BN9PR11MB5276A78546BD2B38068385E18CBA9@BN9PR11MB5276.namprd11.prod.outlook.com> On 6/30/22 4:28 PM, Tian, Kevin wrote: >> From: Lu Baolu <baolu.lu@linux.intel.com> >> Sent: Saturday, June 25, 2022 8:52 PM >> >> +struct iommu_domain_info { >> + struct intel_iommu *iommu; >> + unsigned int refcnt; >> + u16 did; >> +}; >> + >> struct dmar_domain { >> int nid; /* node id */ >> - >> - unsigned int iommu_refcnt[DMAR_UNITS_SUPPORTED]; >> - /* Refcount of devices per iommu */ >> - >> - >> - u16 iommu_did[DMAR_UNITS_SUPPORTED]; >> - /* Domain ids per IOMMU. Use u16 >> since >> - * domain ids are 16 bit wide >> according >> - * to VT-d spec, section 9.3 */ > > It'd make sense to keep the comments when moving above fields. Sure. Updated. diff --git a/drivers/iommu/intel/iommu.h b/drivers/iommu/intel/iommu.h index 56c3d1a9e155..fae45bbb0c7f 100644 --- a/drivers/iommu/intel/iommu.h +++ b/drivers/iommu/intel/iommu.h @@ -527,8 +527,10 @@ struct context_entry { struct iommu_domain_info { struct intel_iommu *iommu; - unsigned int refcnt; - u16 did; + unsigned int refcnt; /* Refcount of devices per iommu */ + u16 did; /* Domain ids per IOMMU. Use u16 since + * domain ids are 16 bit wide according + * to VT-d spec, section 9.3 */ }; > >> + spin_lock(&iommu->lock); >> + curr = xa_load(&domain->iommu_array, iommu->seq_id); >> + if (curr) { >> + curr->refcnt++; >> + kfree(info); >> + goto success; > > Not a fan of adding a label for success. Just putting two lines (unlock+ > return) here is clearer. Updated. > >> + ret = xa_err(xa_store(&domain->iommu_array, iommu->seq_id, >> + info, GFP_ATOMIC)); > > check xa_err in separate line. Sure. Updated as below: @@ -1778,13 +1780,14 @@ static int domain_attach_iommu(struct dmar_domain *domain, info->did = num; info->iommu = iommu; domain->nid = iommu->node; - ret = xa_err(xa_store(&domain->iommu_array, iommu->seq_id, - info, GFP_ATOMIC)); - if (ret) + curr = xa_cmpxchg(&domain->iommu_array, iommu->seq_id, + NULL, info, GFP_ATOMIC); + if (curr) { + ret = xa_err(curr) ? : -EBUSY; goto err_clear; + } domain_update_iommu_cap(domain); > >> >> static void domain_detach_iommu(struct dmar_domain *domain, >> struct intel_iommu *iommu) >> { >> - int num; >> + struct iommu_domain_info *info; >> >> spin_lock(&iommu->lock); >> - domain->iommu_refcnt[iommu->seq_id] -= 1; >> - if (domain->iommu_refcnt[iommu->seq_id] == 0) { >> - num = domain->iommu_did[iommu->seq_id]; >> - clear_bit(num, iommu->domain_ids); >> + info = xa_load(&domain->iommu_array, iommu->seq_id); >> + if (--info->refcnt == 0) { >> + clear_bit(info->did, iommu->domain_ids); >> + xa_erase(&domain->iommu_array, iommu->seq_id); >> domain_update_iommu_cap(domain); >> - domain->iommu_did[iommu->seq_id] = 0; >> + kfree(info); > > domain->nid may still point to the node of the removed iommu. Good catch! I should assign it to NUMA_NO_NODE, so that it could be updated in the next domain_update_iommu_cap(). Updated as below: @@ -1806,6 +1809,7 @@ static void domain_detach_iommu(struct dmar_domain *domain, if (--info->refcnt == 0) { clear_bit(info->did, iommu->domain_ids); xa_erase(&domain->iommu_array, iommu->seq_id); + domain->nid = NUMA_NO_NODE; domain_update_iommu_cap(domain); kfree(info); } Best regards, baolu
WARNING: multiple messages have this Message-ID (diff)
From: Lu Baolu <baolu.lu@linux.intel.com> To: "Tian, Kevin" <kevin.tian@intel.com>, Joerg Roedel <joro@8bytes.org>, Steve Wahl <steve.wahl@hpe.com> Cc: Dimitri Sivanich <sivanich@hpe.com>, "Anderson, Russ" <russ.anderson@hpe.com>, Mike Travis <mike.travis@hpe.com>, Jerry Snitselaar <jsnitsel@redhat.com>, "iommu@lists.linux-foundation.org" <iommu@lists.linux-foundation.org>, "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>, "iommu@lists.linux.dev" <iommu@lists.linux.dev>, David Woodhouse <dwmw2@infradead.org> Subject: Re: [PATCH v1 3/6] iommu/vt-d: Refactor iommu information of each domain Date: Fri, 1 Jul 2022 10:35:50 +0800 [thread overview] Message-ID: <595c314b-2d16-aeef-d808-d1cb713ba175@linux.intel.com> (raw) In-Reply-To: <BN9PR11MB5276A78546BD2B38068385E18CBA9@BN9PR11MB5276.namprd11.prod.outlook.com> On 6/30/22 4:28 PM, Tian, Kevin wrote: >> From: Lu Baolu <baolu.lu@linux.intel.com> >> Sent: Saturday, June 25, 2022 8:52 PM >> >> +struct iommu_domain_info { >> + struct intel_iommu *iommu; >> + unsigned int refcnt; >> + u16 did; >> +}; >> + >> struct dmar_domain { >> int nid; /* node id */ >> - >> - unsigned int iommu_refcnt[DMAR_UNITS_SUPPORTED]; >> - /* Refcount of devices per iommu */ >> - >> - >> - u16 iommu_did[DMAR_UNITS_SUPPORTED]; >> - /* Domain ids per IOMMU. Use u16 >> since >> - * domain ids are 16 bit wide >> according >> - * to VT-d spec, section 9.3 */ > > It'd make sense to keep the comments when moving above fields. Sure. Updated. diff --git a/drivers/iommu/intel/iommu.h b/drivers/iommu/intel/iommu.h index 56c3d1a9e155..fae45bbb0c7f 100644 --- a/drivers/iommu/intel/iommu.h +++ b/drivers/iommu/intel/iommu.h @@ -527,8 +527,10 @@ struct context_entry { struct iommu_domain_info { struct intel_iommu *iommu; - unsigned int refcnt; - u16 did; + unsigned int refcnt; /* Refcount of devices per iommu */ + u16 did; /* Domain ids per IOMMU. Use u16 since + * domain ids are 16 bit wide according + * to VT-d spec, section 9.3 */ }; > >> + spin_lock(&iommu->lock); >> + curr = xa_load(&domain->iommu_array, iommu->seq_id); >> + if (curr) { >> + curr->refcnt++; >> + kfree(info); >> + goto success; > > Not a fan of adding a label for success. Just putting two lines (unlock+ > return) here is clearer. Updated. > >> + ret = xa_err(xa_store(&domain->iommu_array, iommu->seq_id, >> + info, GFP_ATOMIC)); > > check xa_err in separate line. Sure. Updated as below: @@ -1778,13 +1780,14 @@ static int domain_attach_iommu(struct dmar_domain *domain, info->did = num; info->iommu = iommu; domain->nid = iommu->node; - ret = xa_err(xa_store(&domain->iommu_array, iommu->seq_id, - info, GFP_ATOMIC)); - if (ret) + curr = xa_cmpxchg(&domain->iommu_array, iommu->seq_id, + NULL, info, GFP_ATOMIC); + if (curr) { + ret = xa_err(curr) ? : -EBUSY; goto err_clear; + } domain_update_iommu_cap(domain); > >> >> static void domain_detach_iommu(struct dmar_domain *domain, >> struct intel_iommu *iommu) >> { >> - int num; >> + struct iommu_domain_info *info; >> >> spin_lock(&iommu->lock); >> - domain->iommu_refcnt[iommu->seq_id] -= 1; >> - if (domain->iommu_refcnt[iommu->seq_id] == 0) { >> - num = domain->iommu_did[iommu->seq_id]; >> - clear_bit(num, iommu->domain_ids); >> + info = xa_load(&domain->iommu_array, iommu->seq_id); >> + if (--info->refcnt == 0) { >> + clear_bit(info->did, iommu->domain_ids); >> + xa_erase(&domain->iommu_array, iommu->seq_id); >> domain_update_iommu_cap(domain); >> - domain->iommu_did[iommu->seq_id] = 0; >> + kfree(info); > > domain->nid may still point to the node of the removed iommu. Good catch! I should assign it to NUMA_NO_NODE, so that it could be updated in the next domain_update_iommu_cap(). Updated as below: @@ -1806,6 +1809,7 @@ static void domain_detach_iommu(struct dmar_domain *domain, if (--info->refcnt == 0) { clear_bit(info->did, iommu->domain_ids); xa_erase(&domain->iommu_array, iommu->seq_id); + domain->nid = NUMA_NO_NODE; domain_update_iommu_cap(domain); kfree(info); } Best regards, baolu _______________________________________________ iommu mailing list iommu@lists.linux-foundation.org https://lists.linuxfoundation.org/mailman/listinfo/iommu
next prev parent reply other threads:[~2022-07-01 2:40 UTC|newest] Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top 2022-06-25 12:51 [PATCH v1 0/6] iommu/vt-d: Reset DMAR_UNITS_SUPPORTED Lu Baolu 2022-06-25 12:51 ` Lu Baolu 2022-06-25 12:51 ` [PATCH v1 1/6] iommu/vt-d: Remove unused domain_get_iommu() Lu Baolu 2022-06-25 12:51 ` Lu Baolu 2022-06-30 8:19 ` Tian, Kevin 2022-06-30 8:19 ` Tian, Kevin 2022-06-25 12:52 ` [PATCH v1 2/6] iommu/vt-d: Use IDA interface to manage iommu sequence id Lu Baolu 2022-06-25 12:52 ` Lu Baolu 2022-06-30 8:21 ` Tian, Kevin 2022-06-30 8:21 ` Tian, Kevin 2022-07-01 1:17 ` Lu Baolu 2022-07-01 1:17 ` Lu Baolu 2022-06-25 12:52 ` [PATCH v1 3/6] iommu/vt-d: Refactor iommu information of each domain Lu Baolu 2022-06-25 12:52 ` Lu Baolu 2022-06-30 8:28 ` Tian, Kevin 2022-06-30 8:28 ` Tian, Kevin 2022-07-01 2:35 ` Lu Baolu [this message] 2022-07-01 2:35 ` Lu Baolu 2022-06-25 12:52 ` [PATCH v1 4/6] iommu/vt-d: Add VTD_FLAG_IOMMU_PROBED flag Lu Baolu 2022-06-25 12:52 ` Lu Baolu 2022-06-30 8:29 ` Tian, Kevin 2022-06-30 8:29 ` Tian, Kevin 2022-07-01 3:13 ` Baolu Lu 2022-07-01 3:13 ` Baolu Lu 2022-07-01 6:09 ` Tian, Kevin 2022-07-01 6:09 ` Tian, Kevin 2022-06-25 12:52 ` [PATCH v1 5/6] iommu/vt-d: Remove global g_iommus array Lu Baolu 2022-06-25 12:52 ` Lu Baolu 2022-06-30 8:29 ` Tian, Kevin 2022-06-30 8:29 ` Tian, Kevin 2022-06-25 12:52 ` [PATCH v1 6/6] iommu/vt-d: Make DMAR_UNITS_SUPPORTED default 1024 Lu Baolu 2022-06-25 12:52 ` Lu Baolu 2022-06-30 8:30 ` Tian, Kevin 2022-06-30 8:30 ` Tian, Kevin 2022-06-28 12:39 ` [PATCH v1 0/6] iommu/vt-d: Reset DMAR_UNITS_SUPPORTED Baolu Lu 2022-06-28 12:39 ` Baolu Lu 2022-07-06 18:55 ` Steve Wahl
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=595c314b-2d16-aeef-d808-d1cb713ba175@linux.intel.com \ --to=baolu.lu@linux.intel.com \ --cc=dwmw2@infradead.org \ --cc=iommu@lists.linux-foundation.org \ --cc=iommu@lists.linux.dev \ --cc=joro@8bytes.org \ --cc=jsnitsel@redhat.com \ --cc=kevin.tian@intel.com \ --cc=linux-kernel@vger.kernel.org \ --cc=mike.travis@hpe.com \ --cc=russ.anderson@hpe.com \ --cc=sivanich@hpe.com \ --cc=steve.wahl@hpe.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: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.