From: David Hildenbrand <david@redhat.com>
To: Oscar Salvador <osalvador@techadventures.net>, akpm@linux-foundation.org
Cc: mhocko@suse.com, dan.j.williams@intel.com,
yasu.isimatu@gmail.com, rppt@linux.vnet.ibm.com,
malat@debian.org, linux-kernel@vger.kernel.org,
pavel.tatashin@microsoft.com, jglisse@redhat.com,
Jonathan.Cameron@huawei.com, rafael@kernel.org,
dave.jiang@intel.com, linux-mm@kvack.org,
alexander.h.duyck@linux.intel.com,
Oscar Salvador <osalvador@suse.de>
Subject: Re: [PATCH 2/5] mm/memory_hotplug: Create add/del_device_memory functions
Date: Tue, 16 Oct 2018 10:55:56 +0200 [thread overview]
Message-ID: <d0a12eb5-3824-8d25-75f8-3e62f1e81994@redhat.com> (raw)
In-Reply-To: <20181015153034.32203-3-osalvador@techadventures.net>
> index 42d79bcc8aab..d3e52ae71bd9 100644
> --- a/mm/hmm.c
> +++ b/mm/hmm.c
> @@ -996,6 +996,7 @@ static void hmm_devmem_release(struct device *dev, void *data)
> struct zone *zone;
> struct page *page;
> int nid;
> + bool mapping;
>
> if (percpu_ref_tryget_live(&devmem->ref)) {
> dev_WARN(dev, "%s: page mapping is still live!\n", __func__);
> @@ -1010,12 +1011,15 @@ static void hmm_devmem_release(struct device *dev, void *data)
> zone = page_zone(page);
> nid = zone->zone_pgdat->node_id;
>
> - mem_hotplug_begin();
> if (resource->desc == IORES_DESC_DEVICE_PRIVATE_MEMORY)
> - __remove_pages(zone, start_pfn, npages, NULL);
> + mapping = false;
> else
> - arch_remove_memory(nid, start_pfn << PAGE_SHIFT,
> - npages << PAGE_SHIFT, NULL);
> + mapping = true;
> +
> + mem_hotplug_begin();
> + del_device_memory(nid, start_pfn << PAGE_SHIFT, npages << PAGE_SHIFT,
> + NULL,
> + mapping);
> mem_hotplug_done();
>
> hmm_devmem_radix_release(resource);
> @@ -1026,6 +1030,7 @@ static int hmm_devmem_pages_create(struct hmm_devmem *devmem)
> resource_size_t key, align_start, align_size, align_end;
> struct device *device = devmem->device;
> int ret, nid, is_ram;
> + bool mapping;
>
> align_start = devmem->resource->start & ~(PA_SECTION_SIZE - 1);
> align_size = ALIGN(devmem->resource->start +
> @@ -1084,7 +1089,6 @@ static int hmm_devmem_pages_create(struct hmm_devmem *devmem)
> if (nid < 0)
> nid = numa_mem_id();
>
> - mem_hotplug_begin();
> /*
> * For device private memory we call add_pages() as we only need to
> * allocate and initialize struct page for the device memory. More-
> @@ -1096,20 +1100,17 @@ static int hmm_devmem_pages_create(struct hmm_devmem *devmem)
> * want the linear mapping and thus use arch_add_memory().
> */
Some parts of this comment should be moved into add_device_memory now.
(e.g. we call add_pages() ...)
> if (devmem->pagemap.type == MEMORY_DEVICE_PUBLIC)
> - ret = arch_add_memory(nid, align_start, align_size, NULL,
> - false);
> + mapping = true;
> else
> - ret = add_pages(nid, align_start >> PAGE_SHIFT,
> - align_size >> PAGE_SHIFT, NULL, false);
> - if (ret) {
> - mem_hotplug_done();
> - goto error_add_memory;
> - }
> - move_pfn_range_to_zone(&NODE_DATA(nid)->node_zones[ZONE_DEVICE],
> - align_start >> PAGE_SHIFT,
> - align_size >> PAGE_SHIFT, NULL);
> + mapping = false;
> +
> + mem_hotplug_begin();
> + ret = add_device_memory(nid, align_start, align_size, NULL, mapping);
> mem_hotplug_done();
>
> + if (ret)
> + goto error_add_memory;
> +
> /*
> * Initialization of the pages has been deferred until now in order
> * to allow us to do the work while not holding the hotplug lock.
> diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
> index 33d448314b3f..5874aceb81ac 100644
> --- a/mm/memory_hotplug.c
> +++ b/mm/memory_hotplug.c
> @@ -1889,4 +1889,45 @@ void remove_memory(int nid, u64 start, u64 size)
> unlock_device_hotplug();
> }
> EXPORT_SYMBOL_GPL(remove_memory);
> +
> +#ifdef CONFIG_ZONE_DEVICE
> +int del_device_memory(int nid, unsigned long start, unsigned long size,
> + struct vmem_altmap *altmap, bool mapping)
> +{
> + int ret;
nit: personally I prefer short parameters last in the list.
> + unsigned long start_pfn = PHYS_PFN(start);
> + unsigned long nr_pages = size >> PAGE_SHIFT;
> + struct zone *zone = page_zone(pfn_to_page(pfn));
> +
> + if (mapping)
> + ret = arch_remove_memory(nid, start, size, altmap);
> + else
> + ret = __remove_pages(zone, start_pfn, nr_pages, altmap);
> +
> + return ret;
> +}
> +#endif
> #endif /* CONFIG_MEMORY_HOTREMOVE */
> +
> +#ifdef CONFIG_ZONE_DEVICE
> +int add_device_memory(int nid, unsigned long start, unsigned long size,
> + struct vmem_altmap *altmap, bool mapping)
> +{
> + int ret;
dito
> + unsigned long start_pfn = PHYS_PFN(start);
> + unsigned long nr_pages = size >> PAGE_SHIFT;
> +
> + if (mapping)
> + ret = arch_add_memory(nid, start, size, altmap, false);
> + else
> + ret = add_pages(nid, start_pfn, nr_pages, altmap, false);
> +
> + if (!ret) {
> + struct zone *zone = &NODE_DATA(nid)->node_zones[ZONE_DEVICE];
> +
> + move_pfn_range_to_zone(zone, start_pfn, nr_pages, altmap);
> + }
> +
> + return ret;
> +}
> +#endif
>
Can you document for both functions that they should be called with the
memory hotplug lock in write?
Apart from that looks good to me.
--
Thanks,
David / dhildenb
next prev parent reply other threads:[~2018-10-16 8:56 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-10-15 15:30 [PATCH 0/5] Do not touch pages/zones during hot-remove path Oscar Salvador
2018-10-15 15:30 ` [PATCH 1/5] mm/memory_hotplug: Add nid parameter to arch_remove_memory Oscar Salvador
2018-11-12 18:52 ` Pavel Tatashin
2018-10-15 15:30 ` [PATCH 2/5] mm/memory_hotplug: Create add/del_device_memory functions Oscar Salvador
2018-10-16 8:55 ` David Hildenbrand [this message]
2018-10-17 9:33 ` Oscar Salvador
2018-10-17 9:45 ` David Hildenbrand
2018-10-18 6:57 ` Oscar Salvador
2018-11-12 19:14 ` Pavel Tatashin
2018-11-12 19:21 ` Dan Williams
2018-11-12 21:28 ` Pavel Tatashin
2018-11-16 13:06 ` osalvador
2018-11-16 13:03 ` osalvador
2018-10-15 15:30 ` [PATCH 3/5] mm/memory_hotplug: Check for IORESOURCE_SYSRAM in release_mem_region_adjustable Oscar Salvador
2018-10-16 9:54 ` David Hildenbrand
2018-11-12 19:27 ` Pavel Tatashin
2018-10-15 15:30 ` [PATCH 4/5] mm/memory_hotplug: Move zone/pages handling to offline stage Oscar Salvador
2018-10-15 15:30 ` [PATCH 5/5] mm/memory-hotplug: Rework unregister_mem_sect_under_nodes Oscar Salvador
2018-10-18 14:24 ` Jonathan Cameron
2018-10-18 15:02 ` Oscar Salvador
2018-10-18 15:38 ` Jonathan Cameron
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=d0a12eb5-3824-8d25-75f8-3e62f1e81994@redhat.com \
--to=david@redhat.com \
--cc=Jonathan.Cameron@huawei.com \
--cc=akpm@linux-foundation.org \
--cc=alexander.h.duyck@linux.intel.com \
--cc=dan.j.williams@intel.com \
--cc=dave.jiang@intel.com \
--cc=jglisse@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=malat@debian.org \
--cc=mhocko@suse.com \
--cc=osalvador@suse.de \
--cc=osalvador@techadventures.net \
--cc=pavel.tatashin@microsoft.com \
--cc=rafael@kernel.org \
--cc=rppt@linux.vnet.ibm.com \
--cc=yasu.isimatu@gmail.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;
as well as URLs for NNTP newsgroup(s).