Remove unnecessary duplications of "e820_table->entries[i]." via a local
variable, plus pass in 'entry' to the type_to_*() functions which further
improves the readability of the code - and other small tweaks.
No change in functionality.
Cc: Alex Thorlton
Cc: Andy Lutomirski
Cc: Borislav Petkov
The __e820__update_table() function has various weirdly named variables,
such as 'pbios', 'biosmap' and 'pnr_map' which are pretty confusing
and actively misleading at times.
This weird naming found its way into other functions as well, such as
__append_e820_table() and append_e820_table().
Stand
So memblock_x86_fill() is another E820 code misnomer:
- nothing in its name tells us that it's part of the E820 subsystem ...
- The 'fill' wording is ambiguous and doesn't tell us whether it's a single
entry or some process - while the _real_ purpose of the function is hidden,
which is to
E820_X_MAX is defined in a somewhat messy fashion:
- there's a pretty pointless looking #ifndef __KERNEL__ define that
makes no sense in the non-UAPI header anymore,
- part of it is defined in api.h, which is not for type definitions,
- plus it's defined in two headers and the main explana
update_e820() should have 'e820' as a prefix as most of the other E820
functions have - but it's also a bit unclear about its purpose, as
it's unclear what is updated - the whole table, or an entry?
Also, the name does not express that it's a trivial wrapper
around sanitize_e820_table() that also
In line with the rename to 'struct e820_array', harmonize the naming of common
e820
table variable names as well:
e820 => e820_array
e820_saved=> e820_array_saved
e820_map => e820_array
initial_e820 => e820_array_init
This makes the variable names more consistent and
We introduced memblock_find_dma_reserve() in this commit:
6f2a75369e75 x86, memblock: Use
memblock_memory_size()/memblock_free_memory_size() to get correct dma_reserve
But there's several problems with it:
- The changelog is full of typos and is incomprehensible in general, and
the comme
finish_e820_parsing() is closely related to parse_early_params(), but the
name does not tell us this clearly, so rename it to e820__finish_early_params().
Also add a few comments to explain what the function does.
No change in functionality.
Cc: Alex Thorlton
Cc: Andy Lutomirski
Cc: Borislav P
Also do some minor cleanups.
No change in functionality.
Cc: Alex Thorlton
Cc: Andy Lutomirski
Cc: Borislav Petkov
Cc: Brian Gerst
Cc: Dan Williams
Cc: Denys Vlasenko
Cc: H. Peter Anvin
Cc: Huang, Ying
Cc: Josh Poimboeuf
Cc: Juergen Gross
Cc: Linus Torvalds
Cc: Paul Jackson
Cc: Peter
Also clean it up a bit.
No change in functionality.
Cc: Alex Thorlton
Cc: Andy Lutomirski
Cc: Borislav Petkov
Cc: Brian Gerst
Cc: Dan Williams
Cc: Denys Vlasenko
Cc: H. Peter Anvin
Cc: Huang, Ying
Cc: Josh Poimboeuf
Cc: Juergen Gross
Cc: Linus Torvalds
Cc: Paul Jackson
Cc: Peter Zijls
Use a stricter type for struct e820_entry. Add a build-time check to make
sure the compiler won't ever pack the enum into a field smaller than
'int'.
No change in functionality.
Cc: Alex Thorlton
Cc: Andy Lutomirski
Cc: Borislav Petkov
Cc: Brian Gerst
Cc: Dan Williams
Cc: Denys Vlasenko
Cc:
Change e820__mapped_any() and e820__mapped_all()'s return type and
e820__range_remove()'s check_type parameter to bool.
Propagate it into arch/x86/pci/mmconfig-shared.c as this change
affects a function signature there too.
No change in functionality.
Cc: Alex Thorlton
Cc: Andy Lutomirski
Cc:
Use an enum instead of CPP #define.
Also fix various small annoyances in the descriptions of the
various E820 types.
No change in functionality.
Cc: Alex Thorlton
Cc: Andy Lutomirski
Cc: Borislav Petkov
Cc: Brian Gerst
Cc: Dan Williams
Cc: Denys Vlasenko
Cc: H. Peter Anvin
Cc: Huang, Ying
The e820_reserve_setup_data() is local to arch/x86/kernel/setup.c,
but it is E820 functionality - so move it to e820.c to better
isolate E820 functionality.
No change in functionality.
Cc: Alex Thorlton
Cc: Andy Lutomirski
Cc: Borislav Petkov
Cc: Brian Gerst
Cc: Dan Williams
Cc: Denys Vlasen
On Monday 16 January 2017 00:50:57 Finn Thain wrote:
> This series removes some unused code and related comments,
> addresses the warnings generated by 'make W=1' and 'make C=1'
> and fixes a theoretical bug in the bus reset method in atari_scsi.
>
> There's also a patch to add a missing error chec
parse_e820_ext() is very similar to e820__memory_setup_default(), both are
taking bootloader provided data, add it to the E820 table and then
pass it sanitize_e820_table().
Rename it to e820__memory_setup_extended() to better signal their similar role.
No change in functionality.
Cc: Alex Thorlt
The 'e820entry' and 'e820map' names have various annoyances:
- the missing underscore departs from the usual kernel style
and makes the code look weird,
- in the past I kept confusing the 'map' with the 'entry', because
a 'map' is ambiguous in that regard,
- it's not really clear from t
early_reserve_e820() is an early hack for kexec that does a limited fixup of the
mptable and passes it to the kexec kernel as if it was the real thing.
For this it needs to allocate memory - but no memory allocator is available yet
beyond the memblock allocator, so early_reserve_e820() is really a
We've got a number of defines related to the E820 table and its size:
E820MAP
E820NR
E820_X_MAX
E820MAX
The first two denote byte offsets into the zeropage (struct boot_params),
and can are not used in the kernel and can be removed.
The E820_*_MAX values have an i
So there's a number of constants that start with "E820" but which
are not types - these create a confusing mixture when seen together
with 'enum e820_type' values:
E820MAP
E820NR
E820_X_MAX
E820MAX
To better differentiate the 'enum e820_type' values prefix them
wit
Over the last decade or so e820.c has become an ureadable mess of
tinkerware. Perform some very basic cleanups before doing more
intricate cleanups, so that my eyes don't start bleeding when I look at it.
Here's some of the excesses:
- Total disregard of countless aspects of Documentation/Coding
The 'any' and 'all' are modified to the 'mapped' concept, so move them last in
the name.
No change in functionality.
Cc: Alex Thorlton
Cc: Andy Lutomirski
Cc: Borislav Petkov
Cc: Brian Gerst
Cc: Dan Williams
Cc: Denys Vlasenko
Cc: H. Peter Anvin
Cc: Huang, Ying
Cc: Josh Poimboeuf
Cc: Ju
The e820_mark_nosave_regions definition has a number of ugly #ifdef
conditions that unnecessarily uglify both the header and the
e820.c file.
Make this function unconditional: most distro kernels have hibernation
enabled. If LTO functionality is added in the future it will be able
to eliminate unu
bootparam.h, which defines the legacy 'zeropage' boot parameter area,
requires a small amount of e280 defines in the UAPI space - provide them.
No change in functionality.
Cc: Alex Thorlton
Cc: Andy Lutomirski
Cc: Borislav Petkov
Cc: Brian Gerst
Cc: Dan Williams
Cc: Denys Vlasenko
Cc: H. Pe
The e820__update_table() parameters are pretty complex:
arch/x86/include/asm/e820/api.h:extern int e820__update_table(struct
e820_entry *biosmap, int max_nr_map, u32 *pnr_map);
But 90% of the usage is trivial:
arch/x86/kernel/e820.c: if (e820__update_table(e820_table->entries,
ARRAY
Hello
I'm trying (failing) to get iSER working. After rebooting with some settings
saved in targetcli, I got an endless stream of messages like this:
[ 192.701299] isert: isert_setup_id: rdma_bind_addr() failed: -19
[ 192.702733] isert: isert_setup_id: rdma_bind_addr() failed: -19
[ 192.704021
On Sat, Jan 28, 2017 at 2:11 PM, Ingo Molnar wrote:
> Use a stricter type for struct e820_entry. Add a build-time check to make
> sure the compiler won't ever pack the enum into a field smaller than
> 'int'.
I'm not sure this is a good idea. In fact, I'm pretty sure it's a horrible idea.
The com
On Sun, Jan 15, 2017 at 09:03:45PM +0100, Lukas Wunner wrote:
> Hotplug ports generally block their parents from suspending to D3hot as
> otherwise their interrupts couldn't be delivered.
Well, the hotplug ports don't actually block anything. We may decide
not to suspend the parent because normal
Do a number of easy cleanups:
- remove spurious linebreaks
- remove spurious whitespace differences and inconsistent tabulation
- remove unused and ugly 'struct setup_data;' pre-declaration
- make all exported functionality 'extern' consistently
- deobfuscate the (s,e) parameters of is_IS
On 2017-01-27 20:39, Rob Herring wrote:
> On Wed, Jan 18, 2017 at 04:57:10PM +0100, Peter Rosin wrote:
>> Describe how a generic multiplexer controller is used to mux an i2c bus.
>>
>> Acked-by: Jonathan Cameron
>> Signed-off-by: Peter Rosin
>> ---
>> .../devicetree/bindings/i2c/i2c-mux-simple.t
Do some common-sense cleanups:
- standardize on the kernel coding style consistently
- tabulate definitions consistently
- extend and clarify various descriptions
- fix speling
- update the header guard name according to the new position
- etc.
No change in functionality.
Cc: Alex Tho
The E820 region type is put into four different types (!) when used in function
parameters or local variables:
unsigned type;
int type;
unsigned long current_type;
u32 type;
Use 'enum e820_type' in all these cases instead.
No change in functionality.
Cc: Alex Tho
In line with asm/e820/types.h, move the e820 API declarations to
asm/e820/api.h and update all usage sites.
This is just a mechanical, obviously correct move & replace patch,
there will be subsequent changes to clean up the code and to make
better use of the new header organization.
Cc: Alex Thor
There's various sizeof() uses in e820.c - standardize on the shortest
and least error prone one, along the pattern of:
- memset(entry, 0, sizeof(struct e820_entry));
+ memset(entry, 0, sizeof(*entry));
... because with this pattern in most cases it's immediately clear that
we have use
So the E820 code has a lot of cases of:
struct e820_entry *ei;
... but the 'ei' name makes very little sense if you think about it, it's
not an abbreviation of anything obviously related to E820 table entries.
This results in weird looking lines such as:
if (type && ei->t
On Sun, Jan 15, 2017 at 09:03:45PM +0100, Lukas Wunner wrote:
> Since commit 989561de9b51 ("PM / Domains: add setter for dev.pm_domain")
> a PM domain may only be assigned to unbound devices.
>
> The motivation was not made explicit in the changelog other than "in the
> general case that can cause
When reviewing an E820 patch two days ago I noticed that our E820 code
in arch/x86/kernel/e820.c has bitrotten quite a bit, and started
cleaning it up a bit.
Didn't suspect that I'd end up with 50 patches, but oh well ...
The code got the usual treatments to turn it into a modern, cleanly
structu
On Sat, 2017-01-28 at 23:11 +0100, Ingo Molnar wrote:
> No change in functionality.
Not true.
> @@ -419,11 +419,11 @@ __e820_update_range(struct e820_table *table, u64
> start, u64 size, unsigned old_
> size = ULLONG_MAX - start;
>
> end = start + size;
> - printk(KERN_
On Sun, Jan 15, 2017 at 09:03:45PM +0100, Lukas Wunner wrote:
> Document and implement Apple's ACPI-based (but nonstandard) pm mechanism
> for Thunderbolt. Briefly, an ACPI method provided by Apple is used to
> cut power to the controller. A GPE is enabled while the controller is
> powered down w
On Sat, Jan 28, 2017 at 01:44:35PM -0800, Guenter Roeck wrote:
> On 01/28/2017 11:22 AM, Dmitry Torokhov wrote:
> >Guenter, I know you are a coccinelle wizard, can you cook a script that
> >can find current users of clk_enable() in probe paths? Then we can make
> >informed decision on devm_clk_enab
The default_machine_specific_memory_setup() is a mouthful and despite the
many words it doesn't actually tell us clearly what it does.
The function is the x86 legacy memory layout setup code, based on
E820-formatted memory layout information passed by the bootloader
via the boot_params.
Rename it
A commonly used lowlevel x86 header, asm/pgtable.h, includes asm/e820/api.h
spuriously, without making direct use of it.
Removing it is not simple: over the years various .c code learned to rely
on this indirect inclusion.
Remove the unnecessary include - this should speed up the kernel build a b
On Fri, Nov 13, 2015 at 12:46 AM, Satay Epic wrote:
> I've an ARM SL3516 board with Linux 2.6.15 and I've tool-chain setup.
> I could able to compile the GPL source. This board has red-boot.
I have recently started to migrate the Gemini family to device tree.
If you're still working on this bo
CC: triv...@kernel.org
Signed-off-by: Luc Van Oostenryck
---
ipc/mqueue.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/ipc/mqueue.c b/ipc/mqueue.c
index 7a2d8f0c8..4fdd97031 100644
--- a/ipc/mqueue.c
+++ b/ipc/mqueue.c
@@ -558,6 +558,7 @@ static void wq_add(struct mqueue_inode_info *info,
This patch series adds support for the devantech srf04 ultrasonic ranger as
IIO device.
The device is able to recognize objects in a range of 1 cm and 3 meters.
The theoretical resolution is about 3 mm, practically more or less 1 cm.
Use cases for this device are level metering in tanks or distan
This patch adds dt binding for devantech ultrasonic ranger srf04.
The vendor "devantech" was already added to the vendor list with
"[PATCH v4 1/3] iio: distance: srf08: add trivial DT binding"
Signed-off-by: Andreas Klinger
---
.../bindings/iio/proximity/devantech-srf04.txt |
No change in functionality.
Cc: Alex Thorlton
Cc: Andy Lutomirski
Cc: Borislav Petkov
Cc: Brian Gerst
Cc: Dan Williams
Cc: Denys Vlasenko
Cc: H. Peter Anvin
Cc: Huang, Ying
Cc: Josh Poimboeuf
Cc: Juergen Gross
Cc: Linus Torvalds
Cc: Paul Jackson
Cc: Peter Zijlstra
Cc: Rafael J. Wysock
So the 'e820_table_saved' is a bit of a misnomer that hides its real purpose.
At first sight the name suggests that it's some sort save/restore mechanism,
as this is how we typically name such facilities in the kernel.
But that is not so, e820_table_saved is the original firmware version of the
e
This patch adds support for the ultrasonic ranger srf04 of devantech.
This device is measuring the distance of objects in a range between 1 cm
and 3 meters and a theoretical resolution of 3 mm.
There are two GPIOs used:
- trigger: set when the measurement should start
- echo: set when the ult
There's an assembly guard in asm/e820/types.h, and only
a single .S file includes this header: arch/x86/boot/header.S,
but it does not actually make use of any of the E820 defines.
Remove the inclusion and remove the assembly guard as well.
No change in functionality.
Cc: Alex Thorlton
Cc: Andy
asm/e820/api.h had a spurious __ASSEMBLY__ guard - but the
API header is not included in any assembly files. Remove it.
No change in functionality.
Cc: Alex Thorlton
Cc: Andy Lutomirski
Cc: Borislav Petkov
Cc: Brian Gerst
Cc: Dan Williams
Cc: Denys Vlasenko
Cc: H. Peter Anvin
Cc: Huang, Yi
From: wang di
Check if the request is NULL, before retrieve reply body
from the request.
Signed-off-by: wang di
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7396
Reviewed-on: http://review.whamcloud.com/17079
Reviewed-by: John L. Hammond
Reviewed-by: Andreas Dilger
Signed-off-by: James
From: Nathaniel Clark
If a backing filesystem (ZFS) returns that it supports very large
(LLONG_MAX) object sizes, that should be correctly supported. This
fixes the check for unitialized stripe_maxbytes in
lsm_unpackmd_common(), so that ZFS can return LLONG_MAX and it will be
okay. This issue is
From: Ann Koehler
The algorithm for counting freeable objects in the lu_cache shrinker
does not scale with the number of cpus. The LU_SS_LRU_LEN counter
for each cpu is read and summed at shrink time while holding the
lu_sites_guard mutex. With a large number of cpus and low memory
conditions, pr
From: Giuseppe Di Natale
The target_obd debugfs file was not being generated correctly
in cases where nonconsecutive MDT indices were used when
generating a filesystem.
Signed-off-by: Giuseppe Di Natale
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8100
Reviewed-on: http://review.whamclou
From: Niu Yawei
Set mbits for EINPROGRESS resend in ptl_send_rpc().
Signed-off-by: Niu Yawei
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8193
Reviewed-on: http://review.whamcloud.com/20377
Reviewed-by: Liang Zhen
Reviewed-by: Andreas Dilger
Reviewed-by: Oleg Drokin
Signed-off-by: Jam
From: Lai Siyao
This patch dropped support for remote entry statahead, because it
needs 2 async RPCs to fetch both LOOKUP lock from parent MDT and
UPDATE lock from client MDT, which is complicated. Plus not
supporting remote entry statahead won't cause any issue.
* pack child fid in statahead re
From: Jinshan Xiong
ignore_layout can be set for operations that layout won't be changed,
typically page operations. Ignoring layout change in group lock
request will confuse layout change code at LOV layer and hit
assertion.
Signed-off-by: Henri Doreau
Signed-off-by: Jinshan Xiong
Signed-off-
From: Alex Zhuravlev
First of all, this is expensive procedure including a global
mutex and per-bucket spinlocks. also, all the threads observed
exceed will be calling lu_site_purge() and essentially serialized
on that. instead we can let other threads to skip the whole
procedure.
Signed-off-by:
From: frank zago
The function hai_dump_data_field will do a stack buffer
overrun when cat'ing /sys/fs/lustre/.../hsm/actions if an action has
some data in it.
hai_dump_data_field uses snprintf. But there is no check for
truncation, and the value returned by snprintf is used as-is. The
coordinat
From: Ulka Vaze
This issue is found by smatch; has been reported as-
Unchecked usage of potential ERR_PTR result in lmv_hsm_req_count
and lmv_hsm_req_build. Added ERR_PTR in both functions and also
return value check added.
Signed-off-by: Ulka Vaze
Signed-off-by: Aditya Pandit
Intel-bug-id: ht
From: Alexander Boyko
This patch resolves IO vs eviction race.
After eviction failed export stayed at stale list,
a client had IO processing and reconnected during it.
A client sent brw rpc with last lock cookie and new connection.
The lock with failed export was found and assert was happened.
(
The e820_setup_gap() function name is unnecessarily silent about what
kind of gap it sets up. Make it clear that it's about the PCI gap.
No change in functionality.
Cc: Alex Thorlton
Cc: Andy Lutomirski
Cc: Borislav Petkov
Cc: Brian Gerst
Cc: Dan Williams
Cc: Denys Vlasenko
Cc: H. Peter Anv
The HIGH_MEMORY define was in the API header, while it conceptually
belongs to the other physical memory ranges in the e820/types.h
header.
Move it there - and also convert the 1MB address to hexa, so that
it lines up more nicely with the other memory address values.
No change in functionality.
From: "John L. Hammond"
Remove obsolete comments about the behavior of ll_unlink()
Signed-off-by: John L. Hammond
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8003
Reviewed-on: http://review.whamcloud.com/19881
Reviewed-by: Andrew Perepechko
Reviewed-by: James Simmons
Reviewed-by: Oleg
From: "John L. Hammond"
In sptlrpc_gc_add_sec() swap the arguments to list_add_tail() so that
it does what we meant it to do.
Signed-off-by: John L. Hammond
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8270
Reviewed-on: http://review.whamcloud.com/20784
Reviewed-by: Andreas Dilger
Revie
From: Ben Evans
These #if 0 blocks have been in place for years. Assume
they are not used and remove them
Signed-off-by: Ben Evans
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8058
Reviewed-on: http://review.whamcloud.com/20416
Reviewed-by: John L. Hammond
Reviewed-by: Frank Zago
Revie
From: Fan Yong
Originally, the logic of handling config_llog_data::cld_refcount
is some confusing, it may cause the cld_refcount to be leaked or
trigger "LASSERT(atomic_read(&cld->cld_refcount) > 0);" when put
the reference. This patch clean related logic as following:
1) When the 'cld' is creat
From: Oleg Drokin
By forcing creates to always go via lookup we lose some
important caching benefits too.
Instead let's trust creates with positive cached entries.
Then we have 3 possible outcomes:
1. Negative dentry - we go via atomic_open and do the create
by name there.
2. Positive dentry,
From: Fan Yong
There is race condition when allocating fid/seq in parallel
as following:
The thread1 will release the lcs_mutex via seq_fid_alloc_prep(),
then another fid allocation thread2 can obtain the lcs_mutex and
allocate FID in the new sequence that has just been allocated by
the thread1
From: Lai Siyao
Current client doesn't check permission before updating filesystem
default stripe on MGS, which isn't secure and obvious.
Since we setattr on MDS first, and then set default stripe on MGS,
we can just return error upon setattr failure.
Now filesystem default stripe is stored in
From: Andreas Dilger
Change the order of checks for inactive OSCs in lov_prep_statfs_set()
so that administratively disabled OSTs do not generate any output in
"lfs df" at all, to avoid needlessly cluttering the output.
Enable the lazystatfs mount option by default, so that "df" does not
hang wh
From: Bruno Faccini
When a LBUG has occurred, without panic_on_lbug being set,
health_check sysfs file must return an unhealthy state.
Signed-off-by: Bruno Faccini
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7486
Reviewed-on: http://review.whamcloud.com/17981
Reviewed-by: Bobi Jam
Revi
From: Andriy Skulysh
Whole policy structure was zeroed twice. Once during enqueue
and second time during resend or replay. Policy structure
should be initialized with default values only in ldlm_lock_new().
Signed-off-by: Andriy Skulysh
Signed-off-by: Ben Evans
Intel-bug-id: https://jira.hpdd.
From: Bobi Jam
Register cache shrinker to reclaim memory from cl_page LRU list.
Signed-off-by: Bobi Jam
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-6842
Reviewed-on: http://review.whamcloud.com/15630
Reviewed-by: Andreas Dilger
Reviewed-by: Jinshan Xiong
Reviewed-by: Oleg Drokin
Sign
From: Dmitry Eremin
Setting extended attributes permissions are properly checked with and
without ACLs. In user.* namespace, only regular files and directories
can have extended attributes. For sticky directories, only the owner
and privileged user can write attributes.
Signed-off-by: Dmitry Ere
From: Fan Yong
The 'fld_read_server' uses 'RMF_GENERIC_DATA' to hold the 'FLD_QUERY'
RPC reply that is composed of 'struct lu_seq_range_array'. But there
is not registered swabber function for 'RMF_GENERIC_DATA'. So the RPC
peers need to handle the RPC reply with fixed little-endian format.
In t
From: Jeremy Filizetti
Restore connect flags on failure of ptlrpc_connect_import()
to prevent an LBUG due to flags mismatch.
Signed-off-by: Jeremy Filizetti
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7185
Reviewed-on: http://review.whamcloud.com/16950
Reviewed-by: Andreas Dilger
Revie
From: Andreas Dilger
Update the sysfs "version" file to print "lustre: " with
the version number.
Signed-off-by: Andreas Dilger
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5969
Reviewed-on: http://review.whamcloud.com/16721
Reviewed-by: James Simmons
Reviewed-by: Dmitry Eremin
Reviewe
From: Niu Yawei
Sync write should update m/ctime promptly, otherwise, stale m/ctime
could be updated on the OST object by the sync write RPC.
Signed-off-by: Niu Yawei
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7310
Reviewed-on: http://review.whamcloud.com/21063
Reviewed-by: John L. Ham
Batch of missing fixes for lustre for the upstream client.
Alex Zhuravlev (1):
staging: lustre: obdclass: do not call lu_site_purge() for single object
exceed
Alexander Boyko (1):
staging: lustre: ptlrpc: skip lock if export failed
Andreas Dilger (3):
staging: lustre: mdc: quiet console m
From: Jinshan Xiong
Read ahead currently doesn't handle 16MB RPC packets correctly
by assuming the packets are a default size instead of querying
the size. This work adjust the read ahead policy to issue
read ahead RPC by the underlying RPC size.
Signed-off-by: Jinshan Xiong
Signed-off-by: Gu Z
From: Dmitry Eremin
The function cfs_cpt_table_create_pattern() alters the string
passed to it. Currently we are passing in the module parameter
string cpu_pattern which is incorrect. Instead lets duplicate
the module parameter string and pass that to the function
cfs_cpt_table_create_pattern().
From: "John L. Hammond"
Remove the unused lmv.*.placement parameter along with supporting
functions and struct members.
Signed-off-by: John L. Hammond
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7674
Reviewed-on: http://review.whamcloud.com/18019
Reviewed-by: Ben Evans
Reviewed-by: Fra
From: Jinshan Xiong
OSC has to make sure that it won't issue write RPCs with too many
chunks otherwise it will casue ZFS to create transactions much
bigger than DMU_MAX_ACCESS in size, which will end up with write
failure.
Signed-off-by: Jinshan Xiong
Signed-off-by: Dmitry Eremin
Intel-bug-id:
From: Andreas Dilger
If a user process is waiting for MDS recovery during close, but the
process is interrupted, the file is still closed but it prints a
message on the console. Quiet the console message for -EINTR, since
this is expected behaviour.
Signed-off-by: Andreas Dilger
Intel-bug-id: h
In the case of interval_tree.h only interval_set()
uses LASSERT which is removed in this patch and
interval_set() instead reports a real error. The
header libcfs.h for interval_tree.h is not needed
anymore so we can just use the standard linux
kernel headers instead.h
Signed-off-by: James Simmons
From: Bobi Jam
lov_device::ld_target[ost_idx] could be NULL if the OST target is
not filled in lov_device::ld_lov::lov_tgt_desc[ost_idx] yet.
Signed-off-by: Bobi Jam
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8018
Reviewed-on: http://review.whamcloud.com/21411
Reviewed-by: Jinshan Xion
From: Dmitry Eremin
Change default value of CPT pattern and make it match NUMA topology
Signed-off-by: Liang Zhen
Signed-off-by: Dmitry Eremin
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5050
Reviewed-on: http://review.whamcloud.com/22377
Reviewed-by: James Simmons
Reviewed-by: Olaf W
From: Bobi Jam
So that debug log only contains relevant messages for debugging
purpose.
Signed-off-by: Bobi Jam
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8413
Reviewed-on: http://review.whamcloud.com/22753
Reviewed-by: Andreas Dilger
Reviewed-by: Fan Yong
Reviewed-by: Oleg Drokin
S
Replace the ulong_ptr_t and long_ptr_t with standard
kernel types.
Signed-off-by: James Simmons
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-6245
Reviewed-on: http://review.whamcloud.com/20204
Reviewed-by: Frank Zago
Reviewed-by: Dmitry Eremin
Reviewed-by: Oleg Drokin
Signed-off-by: Jam
From: Vladimir Saveliev
ptlrpc_import_delay_req() refuses to delay blocking asts when import
is not in LUSTRE_IMP_FULL yet. That leads to client eviction assuming
that it failed to respond.
Allow delays for blocking asts being resent.
Signed-off-by: Vladimir Saveliev
Intel-bug-id: https://jira
From: Steve Guminski
This patch makes no functional changes. Struct initializers in the
libcfs directory that use C89 or GCC-only syntax are updated to C99
syntax.
The C99 syntax prevents incorrect initialization if values are
accidently placed in the wrong position, allows changes in the struct
From: Steve Guminski
The patch updates the prototype in osc_internal.h to match the
enums used in the declaration.
The osc_match_base declaration in lustre/osc/osc_request.c uses
enums for stricter checking on the type and mode parameters:
int osc_match_base(struct obd_export *exp,
From: "John L. Hammond"
None of the obd_notify() handlers listen for the OBD_NOTIFY_CREATE
event, so remove it and its sole use in lov_add_target().
Signed-off-by: John L. Hammond
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8403
Reviewed-on: https://review.whamcloud.com/21420
Reviewed-b
Remove the inline function socklnd_init_msg.
Its only used by the kernel code so no point
keeping it in an UAPI header.
Signed-off-by: James Simmons
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-6142
Reviewed-on: https://review.whamcloud.com/18506
Reviewed-by: Dmitry Eremin
Reviewed-by: Do
From: Liang Zhen
ksocklnd reaper thread always tries to close the connection for the
first timedout zero-copy TX. This is wrong if this connection is
already being closed, because the reaper will see the same TX again
and again and cannot find out other timedout zero-copy TXs and close
connection
From: Dmitry Eremin
Update max_ptlrpcds module parameter descriptions to let
users know its obsolete. Change cpt to CPT for the module
parameter description ptlrpcd_per_cpt_max so it matches
documentation.
Signed-off-by: Dmitry Eremin
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8890
Rev
From: Niu Yawei
The replay cursor should be updated properly when close happened
during replay, otherwise, ptlrpc_replay_next() could run into a
dead loop due to an invalid replay cursor:
- replay cursor is moved to an open request during replay;
- application close that open file, so the rq_rep
From: Dmitry Eremin
Environment for request interpreters is not used any more.
Signed-off-by: Dmitry Eremin
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8887
Reviewed-on: https://review.whamcloud.com/24061
Reviewed-by: John L. Hammond
Reviewed-by: Bob Glossman
Reviewed-by: Oleg Drokin
201 - 300 of 330 matches
Mail list logo
