[PATCH 0/2] mm: devmap: page-freeing related cleanups
by John Hubbard
Hi,
Here are the small devmap page-freeing cleanups as requested, so that
they can be reviewed separately from the larger "track dma-pinned pages
series" [1] that they were originally part of.
I did not add Jan Kara's and Jerome Glisse's reviewed-by tags, because those
were for patch 2 but as part of [1], which is different enough from this
series that I couldn't treat it as "close enough".
Testing: Dan notes that his patch (1) has passed nvdimm unit tests. However,
I have not yet done much testing of the two patches together. What
I've done so far is to boot up a system that has CONFIG_DEV_PAGEMAP_OPS=y
and run LTP tests on it, but none of those tests are exercising
free_devmap_managed_page().
Dan, if you could run this combined set through your unit tests, I'd feel
a lot better about it. I'll also work on getting my HMM test setup revived.
[1] https://lore.kernel.org/r/20191113042710.3997854-1-jhubbard@nvidia.com
Dan Williams (1):
mm: Cleanup __put_devmap_managed_page() vs ->page_free()
John Hubbard (1):
mm: devmap: refactor 1-based refcounting for ZONE_DEVICE pages
drivers/nvdimm/pmem.c | 6 ----
include/linux/mm.h | 27 +++++++++++++--
mm/memremap.c | 76 ++++++++++++++++++++-----------------------
3 files changed, 60 insertions(+), 49 deletions(-)
--
2.24.0
1 year, 2 months
Re: GOODNEWS YOUR FUND IS RELEASED.
by THE IMF AGENT - SWITZERLAND
Reply To;; imf.agent(a)aol.com
Dear Beneficiary,
IMF FUND COMPENSATION AND PAYMENT IN RELEASE OF LOTTERY, UNSETTLED CONTRACTS, LAWYERS & OTHER UNPAID-FUNDS.
Your compensation fund is ready to be received to your would be nominated country and account which is $4,000,000 Our organization confirmed this update today. But we need the info as stated below from you to enable us submit it to the designated paying bank which will be revealed to you upon confirmation of the required details to match-up your email address in our database. This fund is deposited in your favor by The IMF in Geneva Switzerland which submitted your email in respect to your current predicament. It is very important that you reply this message immediately you receive it.
NOTE: We ADVICE that you STOP contacting the Non and not competent Lawyers, Finance houses, Banks and Bankers and so called Government officials claiming to be have your fund in their custody. That is not true anymore because we THE IMF have instructed our licensed partner bank to make payment of $4 million US Dollars to you.
Once again kindly re-confirm your info.
1. NAMES:
2. CONTACT ADDRESS:
3. NATIONALITY/ORIGIN:
4. CURRENT COUNTRY:
5. DATE OF BIRTH:
6. MAN/WOMAN:
7. JOB TYPE:
8. MOBILE NUMBER:
9. EMAIL ADDRESS:
10. SINGLE/MARRIED
All info should be sent to imf.agent(a)aol.com
Congratulation.s and have a nice day!!!
Joseph K. Ingram
Senior Director
For; IMF Geneva-Switzerland.
Email To: imf.agent(a)aol.com
1 year, 2 months
[PATCH v5] libnvdimm/namespace: Differentiate between probe mapping
and runtime mapping
by Dan Williams
From: Aneesh Kumar K.V <aneesh.kumar(a)linux.ibm.com>
The nvdimm core currently maps the full namespace to an ioremap range
while probing the namespace mode. This can result in probe failures on
architectures that have limited ioremap space.
For example, with a large btt namespace that consumes most of I/O remap
range, depending on the sequence of namespace initialization, the user
can find a pfn namespace initialization failure due to unavailable I/O
remap space which nvdimm core uses for temporary mapping.
nvdimm core can avoid this failure by only mapping the reserved info
block area to check for pfn superblock type and map the full namespace
resource only before using the namespace.
Given that personalities like BTT can be layered on top of any namespace
type create a generic form of devm_nsio_enable (devm_namespace_enable)
and use it inside the per-personality attach routines. Now
devm_namespace_enable() is always paired with disable unless the mapping
is going to be used for long term runtime access.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar(a)linux.ibm.com>
Link: https://lore.kernel.org/r/20191017073308.32645-1-aneesh.kumar@linux.ibm.com
[djbw: reworks to move devm_namespace_{en,dis}able into *attach helpers]
Reported-by: kbuild test robot <lkp(a)intel.com>
Link: https://lore.kernel.org/r/20191031105741.102793-2-aneesh.kumar@linux.ibm.com
Signed-off-by: Dan Williams <dan.j.williams(a)intel.com>
---
Changes since v4:
- Fix a 0day report for devm_nsio_disable() being redefined due to it
not being inlined.
drivers/dax/pmem/core.c | 6 +++---
drivers/nvdimm/btt.c | 10 ++++++++--
drivers/nvdimm/claim.c | 14 ++++++--------
drivers/nvdimm/namespace_devs.c | 17 +++++++++++++++++
drivers/nvdimm/nd-core.h | 17 +++++++++++++++++
drivers/nvdimm/nd.h | 22 +++++++++-------------
drivers/nvdimm/pfn_devs.c | 18 +++++++++++-------
drivers/nvdimm/pmem.c | 17 +++++++++++++----
8 files changed, 84 insertions(+), 37 deletions(-)
diff --git a/drivers/dax/pmem/core.c b/drivers/dax/pmem/core.c
index 6eb6dfdf19bf..2bedf8414fff 100644
--- a/drivers/dax/pmem/core.c
+++ b/drivers/dax/pmem/core.c
@@ -25,20 +25,20 @@ struct dev_dax *__dax_pmem_probe(struct device *dev, enum dev_dax_subsys subsys)
ndns = nvdimm_namespace_common_probe(dev);
if (IS_ERR(ndns))
return ERR_CAST(ndns);
- nsio = to_nd_namespace_io(&ndns->dev);
/* parse the 'pfn' info block via ->rw_bytes */
- rc = devm_nsio_enable(dev, nsio);
+ rc = devm_namespace_enable(dev, ndns, nd_info_block_reserve());
if (rc)
return ERR_PTR(rc);
rc = nvdimm_setup_pfn(nd_pfn, &pgmap);
if (rc)
return ERR_PTR(rc);
- devm_nsio_disable(dev, nsio);
+ devm_namespace_disable(dev, ndns);
/* reserve the metadata area, device-dax will reserve the data */
pfn_sb = nd_pfn->pfn_sb;
offset = le64_to_cpu(pfn_sb->dataoff);
+ nsio = to_nd_namespace_io(&ndns->dev);
if (!devm_request_mem_region(dev, nsio->res.start, offset,
dev_name(&ndns->dev))) {
dev_warn(dev, "could not reserve metadata\n");
diff --git a/drivers/nvdimm/btt.c b/drivers/nvdimm/btt.c
index 3e9f45aec8d1..8cb890a987b0 100644
--- a/drivers/nvdimm/btt.c
+++ b/drivers/nvdimm/btt.c
@@ -1674,7 +1674,8 @@ int nvdimm_namespace_attach_btt(struct nd_namespace_common *ndns)
struct nd_region *nd_region;
struct btt_sb *btt_sb;
struct btt *btt;
- size_t rawsize;
+ size_t size, rawsize;
+ int rc;
if (!nd_btt->uuid || !nd_btt->ndns || !nd_btt->lbasize) {
dev_dbg(&nd_btt->dev, "incomplete btt configuration\n");
@@ -1685,6 +1686,11 @@ int nvdimm_namespace_attach_btt(struct nd_namespace_common *ndns)
if (!btt_sb)
return -ENOMEM;
+ size = nvdimm_namespace_capacity(ndns);
+ rc = devm_namespace_enable(&nd_btt->dev, ndns, size);
+ if (rc)
+ return rc;
+
/*
* If this returns < 0, that is ok as it just means there wasn't
* an existing BTT, and we're creating a new one. We still need to
@@ -1693,7 +1699,7 @@ int nvdimm_namespace_attach_btt(struct nd_namespace_common *ndns)
*/
nd_btt_version(nd_btt, ndns, btt_sb);
- rawsize = nvdimm_namespace_capacity(ndns) - nd_btt->initial_offset;
+ rawsize = size - nd_btt->initial_offset;
if (rawsize < ARENA_MIN_SIZE) {
dev_dbg(&nd_btt->dev, "%s must be at least %ld bytes\n",
dev_name(&ndns->dev),
diff --git a/drivers/nvdimm/claim.c b/drivers/nvdimm/claim.c
index 2985ca949912..45964acba944 100644
--- a/drivers/nvdimm/claim.c
+++ b/drivers/nvdimm/claim.c
@@ -300,13 +300,14 @@ static int nsio_rw_bytes(struct nd_namespace_common *ndns,
return rc;
}
-int devm_nsio_enable(struct device *dev, struct nd_namespace_io *nsio)
+int devm_nsio_enable(struct device *dev, struct nd_namespace_io *nsio,
+ resource_size_t size)
{
struct resource *res = &nsio->res;
struct nd_namespace_common *ndns = &nsio->common;
- nsio->size = resource_size(res);
- if (!devm_request_mem_region(dev, res->start, resource_size(res),
+ nsio->size = size;
+ if (!devm_request_mem_region(dev, res->start, size,
dev_name(&ndns->dev))) {
dev_warn(dev, "could not reserve region %pR\n", res);
return -EBUSY;
@@ -318,12 +319,10 @@ int devm_nsio_enable(struct device *dev, struct nd_namespace_io *nsio)
nvdimm_badblocks_populate(to_nd_region(ndns->dev.parent), &nsio->bb,
&nsio->res);
- nsio->addr = devm_memremap(dev, res->start, resource_size(res),
- ARCH_MEMREMAP_PMEM);
+ nsio->addr = devm_memremap(dev, res->start, size, ARCH_MEMREMAP_PMEM);
return PTR_ERR_OR_ZERO(nsio->addr);
}
-EXPORT_SYMBOL_GPL(devm_nsio_enable);
void devm_nsio_disable(struct device *dev, struct nd_namespace_io *nsio)
{
@@ -331,6 +330,5 @@ void devm_nsio_disable(struct device *dev, struct nd_namespace_io *nsio)
devm_memunmap(dev, nsio->addr);
devm_exit_badblocks(dev, &nsio->bb);
- devm_release_mem_region(dev, res->start, resource_size(res));
+ devm_release_mem_region(dev, res->start, nsio->size);
}
-EXPORT_SYMBOL_GPL(devm_nsio_disable);
diff --git a/drivers/nvdimm/namespace_devs.c b/drivers/nvdimm/namespace_devs.c
index 829f7bdaaa6d..c90664387ee5 100644
--- a/drivers/nvdimm/namespace_devs.c
+++ b/drivers/nvdimm/namespace_devs.c
@@ -1759,6 +1759,23 @@ struct nd_namespace_common *nvdimm_namespace_common_probe(struct device *dev)
}
EXPORT_SYMBOL(nvdimm_namespace_common_probe);
+int devm_namespace_enable(struct device *dev, struct nd_namespace_common *ndns,
+ resource_size_t size)
+{
+ if (is_namespace_blk(&ndns->dev))
+ return 0;
+ return devm_nsio_enable(dev, to_nd_namespace_io(&ndns->dev), size);
+}
+EXPORT_SYMBOL_GPL(devm_namespace_enable);
+
+void devm_namespace_disable(struct device *dev, struct nd_namespace_common *ndns)
+{
+ if (is_namespace_blk(&ndns->dev))
+ return;
+ devm_nsio_disable(dev, to_nd_namespace_io(&ndns->dev));
+}
+EXPORT_SYMBOL_GPL(devm_namespace_disable);
+
static struct device **create_namespace_io(struct nd_region *nd_region)
{
struct nd_namespace_io *nsio;
diff --git a/drivers/nvdimm/nd-core.h b/drivers/nvdimm/nd-core.h
index 25fa121104d0..96e8630f451c 100644
--- a/drivers/nvdimm/nd-core.h
+++ b/drivers/nvdimm/nd-core.h
@@ -171,6 +171,23 @@ ssize_t nd_namespace_store(struct device *dev,
struct nd_pfn *to_nd_pfn_safe(struct device *dev);
bool is_nvdimm_bus(struct device *dev);
+#if IS_ENABLED(CONFIG_ND_CLAIM)
+int devm_nsio_enable(struct device *dev, struct nd_namespace_io *nsio,
+ resource_size_t size);
+void devm_nsio_disable(struct device *dev, struct nd_namespace_io *nsio);
+#else
+static inline int devm_nsio_enable(struct device *dev,
+ struct nd_namespace_io *nsio, resource_size_t size)
+{
+ return -ENXIO;
+}
+
+static inline void devm_nsio_disable(struct device *dev,
+ struct nd_namespace_io *nsio)
+{
+}
+#endif
+
#ifdef CONFIG_PROVE_LOCKING
extern struct class *nd_class;
diff --git a/drivers/nvdimm/nd.h b/drivers/nvdimm/nd.h
index ee5c04070ef9..a9f338d01a55 100644
--- a/drivers/nvdimm/nd.h
+++ b/drivers/nvdimm/nd.h
@@ -212,6 +212,11 @@ struct nd_dax {
struct nd_pfn nd_pfn;
};
+static inline u32 nd_info_block_reserve(void)
+{
+ return ALIGN(SZ_8K, PAGE_SIZE);
+}
+
enum nd_async_mode {
ND_SYNC,
ND_ASYNC,
@@ -370,29 +375,20 @@ const char *nvdimm_namespace_disk_name(struct nd_namespace_common *ndns,
unsigned int pmem_sector_size(struct nd_namespace_common *ndns);
void nvdimm_badblocks_populate(struct nd_region *nd_region,
struct badblocks *bb, const struct resource *res);
+int devm_namespace_enable(struct device *dev, struct nd_namespace_common *ndns,
+ resource_size_t size);
+void devm_namespace_disable(struct device *dev,
+ struct nd_namespace_common *ndns);
#if IS_ENABLED(CONFIG_ND_CLAIM)
-
/* max struct page size independent of kernel config */
#define MAX_STRUCT_PAGE_SIZE 64
-
int nvdimm_setup_pfn(struct nd_pfn *nd_pfn, struct dev_pagemap *pgmap);
-int devm_nsio_enable(struct device *dev, struct nd_namespace_io *nsio);
-void devm_nsio_disable(struct device *dev, struct nd_namespace_io *nsio);
#else
static inline int nvdimm_setup_pfn(struct nd_pfn *nd_pfn,
struct dev_pagemap *pgmap)
{
return -ENXIO;
}
-static inline int devm_nsio_enable(struct device *dev,
- struct nd_namespace_io *nsio)
-{
- return -ENXIO;
-}
-static inline void devm_nsio_disable(struct device *dev,
- struct nd_namespace_io *nsio)
-{
-}
#endif
int nd_blk_region_init(struct nd_region *nd_region);
int nd_region_activate(struct nd_region *nd_region);
diff --git a/drivers/nvdimm/pfn_devs.c b/drivers/nvdimm/pfn_devs.c
index 96727fd493f7..3ca6c97cd14d 100644
--- a/drivers/nvdimm/pfn_devs.c
+++ b/drivers/nvdimm/pfn_devs.c
@@ -382,6 +382,15 @@ static int nd_pfn_clear_memmap_errors(struct nd_pfn *nd_pfn)
meta_start = (SZ_4K + sizeof(*pfn_sb)) >> 9;
meta_num = (le64_to_cpu(pfn_sb->dataoff) >> 9) - meta_start;
+ /*
+ * re-enable the namespace with correct size so that we can access
+ * the device memmap area.
+ */
+ devm_namespace_disable(&nd_pfn->dev, ndns);
+ rc = devm_namespace_enable(&nd_pfn->dev, ndns, le64_to_cpu(pfn_sb->dataoff));
+ if (rc)
+ return rc;
+
do {
unsigned long zero_len;
u64 nsoff;
@@ -635,11 +644,6 @@ int nd_pfn_probe(struct device *dev, struct nd_namespace_common *ndns)
}
EXPORT_SYMBOL(nd_pfn_probe);
-static u32 info_block_reserve(void)
-{
- return ALIGN(SZ_8K, PAGE_SIZE);
-}
-
/*
* We hotplug memory at sub-section granularity, pad the reserved area
* from the previous section base to the namespace base address.
@@ -653,7 +657,7 @@ static unsigned long init_altmap_base(resource_size_t base)
static unsigned long init_altmap_reserve(resource_size_t base)
{
- unsigned long reserve = info_block_reserve() >> PAGE_SHIFT;
+ unsigned long reserve = nd_info_block_reserve() >> PAGE_SHIFT;
unsigned long base_pfn = PHYS_PFN(base);
reserve += base_pfn - SUBSECTION_ALIGN_DOWN(base_pfn);
@@ -668,7 +672,7 @@ static int __nvdimm_setup_pfn(struct nd_pfn *nd_pfn, struct dev_pagemap *pgmap)
u64 offset = le64_to_cpu(pfn_sb->dataoff);
u32 start_pad = __le32_to_cpu(pfn_sb->start_pad);
u32 end_trunc = __le32_to_cpu(pfn_sb->end_trunc);
- u32 reserve = info_block_reserve();
+ u32 reserve = nd_info_block_reserve();
struct nd_namespace_common *ndns = nd_pfn->ndns;
struct nd_namespace_io *nsio = to_nd_namespace_io(&ndns->dev);
resource_size_t base = nsio->res.start + start_pad;
diff --git a/drivers/nvdimm/pmem.c b/drivers/nvdimm/pmem.c
index f9f76f6ba07b..7a6f4501dcda 100644
--- a/drivers/nvdimm/pmem.c
+++ b/drivers/nvdimm/pmem.c
@@ -372,6 +372,10 @@ static int pmem_attach_disk(struct device *dev,
if (!pmem)
return -ENOMEM;
+ rc = devm_namespace_enable(dev, ndns, nd_info_block_reserve());
+ if (rc)
+ return rc;
+
/* while nsio_rw_bytes is active, parse a pfn info block if present */
if (is_nd_pfn(dev)) {
nd_pfn = to_nd_pfn(dev);
@@ -381,7 +385,7 @@ static int pmem_attach_disk(struct device *dev,
}
/* we're attaching a block device, disable raw namespace access */
- devm_nsio_disable(dev, nsio);
+ devm_namespace_disable(dev, ndns);
dev_set_drvdata(dev, pmem);
pmem->phys_addr = res->start;
@@ -497,15 +501,16 @@ static int nd_pmem_probe(struct device *dev)
if (IS_ERR(ndns))
return PTR_ERR(ndns);
- if (devm_nsio_enable(dev, to_nd_namespace_io(&ndns->dev)))
- return -ENXIO;
-
if (is_nd_btt(dev))
return nvdimm_namespace_attach_btt(ndns);
if (is_nd_pfn(dev))
return pmem_attach_disk(dev, ndns);
+ ret = devm_namespace_enable(dev, ndns, nd_info_block_reserve());
+ if (ret)
+ return ret;
+
ret = nd_btt_probe(dev, ndns);
if (ret == 0)
return -ENXIO;
@@ -532,6 +537,10 @@ static int nd_pmem_probe(struct device *dev)
return -ENXIO;
else if (ret == -EOPNOTSUPP)
return ret;
+
+ /* probe complete, attach handles namespace enabling */
+ devm_namespace_disable(dev, ndns);
+
return pmem_attach_disk(dev, ndns);
}
1 year, 2 months
Стратегия продаж В2В
by Bronislaw
Профессиональные продажи В2В
Пройдет: 28 ноября, Киев.
Цель программы:
- выстроить адаптированную под специфику бизнеса стратегию привлечения и сопровождения клиента;
- понять концепцию "воронки продаж", научиться системно определять потенциал клиента, видеть и разрабатывать возможности для сделки, ориентироваться на проработку всей клиентской базы, а не только "легких" клиентов, планировать долгосрочные цели работы с клиентами;
- внедрить практику постоянного комплексного контакта с клиентом, позволяющего реализовывать потенциал, возможности для сделки и строить долгосрочные отношения с клиентом;
- эффективно собирать информацию и устанавливать контакт с потенциальными клиентами;
- точно выявлять интересы, потребности и выгоды различных "агентов влияния" в Компании-Заказчика;
- формировать персонализированные коммерческие предложения;
- успешно вести переговоры по телефону и при встрече;
- использовать возражения и сомнения клиента, как возможность для продажи;
- проводить эффективные презентации продукта/услуги в зависимости от характеристик товара и психологического типа клиента;
- побуждать клиента к совершению покупки и доводить клиента до завершения продажи/заключения сделки;
- понимать важность послепродажного сервиса и повторной продажи.
Ознакомиться с программой>
Занятия проводит:
Дегтярева Наталья - с опытом публичных выступлений, обучения и развития групп. Среди клиентов: Издательство "Стандарт", "Oriflame", "Gillette Ukraine LLC", "Новые окна", "Daewoo electronics", "Белла-Трейд", "Каркас", "Кредитпромбанк", "Faberlic", "Райффайзенбанк", "Mary Kay", "Ланит", "Коммункомплекс", "Магнат", "21 век", "Дила", "Найхром", "Форум", "Содружество" и др.
Стоимость участия:
3,000.00 грн. - за одного участника.
Для второго и третьего участника скидки 5% и 7% соответственно
Ознакомиться с программой> http://sales-b2b.in.ua/training/148/professionalnie-prodazhi-v2v-ot-pervo...
Если у Вас возникнут дополнительные вопросы - обращайтесь, мы всегда рады Вам помочь!
--
С уважением,
Bronislaw
email-маркетолог
Больше не получать.
List-Unsubscribe from the newsletter оr complain аbоut SPАМ.
1 year, 2 months
Goodnews, I have deposited your transfer total amount US$4.8million
Dollars with Money Gram this morning. we agreed you will be receiving it
$5000.00 daily.
by Ms.Mary Coster
Attn, Dear
Goodnews, I have deposited your transfer total amount US$4.8million
Dollars with Money Gram this morning. we agreed you will be receiving
it $5000.00 daily.
Contact Mr. John Dave Director, Money Gram to pick up your first Money
Gram payment $5000.00 today.
Contact Person; Mr. John Dave Director, Money Gram,International
Remittance-Benin
Email; moneygram.1820(a)outlook.fr
Telephone; +229 62619517
Please re-confirm your address to him once again such as listed below.
1.Your Full Name..............................
2.Address.........................
3.Country....................
4.Sex.........................................
5.Your telephone numbers..........................
6. Copy of your ID...........................
This is to avoid sending your funds to wrong person, He is waiting to
hear from you urgent today.
Let me know once you pick up your transfer $5000.00 today.
Finally, Note I have paid for the service fees, but only money will
send to him is $90.00 transfer fee before you can pick up the transfer
today.
Ask, Mr. John Dave Director, Money Gram to give you direction where to
send your transfer fee $90.00 only to Him Immediately so that you can
pick up $5000.00 us dollars today.
Thanks for undrstanding.
Mary Coster
m.coster(a)aol.com
1 year, 2 months
Complete PO for confirmation
by info＠primeimports.ga
sorry for contacting you via our other email address
for your confirmation., please find PO copy as attached so we can
start shipment of items.
-----
Regards
PRIMEIMPORTS LTD
Purchase Department
info(a)primeimports.ga
+18263825568
1 year, 2 months
