> On 26 Jun 2018, at 21.01, Heiner Litz wrote:
>
>> Guarantees that this is at least 32bit, but this should be at least 64
>> bit right?
>
> All 64-bit bitmaps in pblk are unsigned long. If we want to change to
> u64 this should be part of a different patch
It makes sense to me.
Javier
> On 26 Jun 2018, at 20.47, Heiner Litz wrote:
>
>> On Fri, Jun 22, 2018 at 11:17 AM Matias Bjørling wrote:
>>
>>> On 06/18/2018 07:56 PM, Heiner Litz wrote:
>>> In the read path, partial reads are currently performed synchronously
>>> which affects performance for workloads that generate
> Guarantees that this is at least 32bit, but this should be at least 64
> bit right?
All 64-bit bitmaps in pblk are unsigned long. If we want to change to
u64 this should be part of a different patch
On Fri, Jun 22, 2018 at 11:17 AM Matias Bjørling wrote:
>
> On 06/18/2018 07:56 PM, Heiner
On Fri, Jun 22, 2018 at 11:17 AM Matias Bjørling wrote:
>
> On 06/18/2018 07:56 PM, Heiner Litz wrote:
> > In the read path, partial reads are currently performed synchronously
> > which affects performance for workloads that generate many partial
> > reads. This patch adds an asynchronous
On 6/22/18 12:17 PM, Matias Bjørling wrote:
>> +/* Re-use allocated memory for intermediate lbas */
>> +lba_list_mem = (((void *)rqd->ppa_list) + pblk_dma_ppa_size);
>> +
>> +new_bio = bio_alloc(GFP_KERNEL, nr_holes);
>
>
> new_bio can return NULL.
It can't, not if __GFP_WAIT is set
On 06/18/2018 07:56 PM, Heiner Litz wrote:
In the read path, partial reads are currently performed synchronously
which affects performance for workloads that generate many partial
reads. This patch adds an asynchronous partial read path as well as
the required partial read ctx.
Signed-off-by:
In the read path, partial reads are currently performed synchronously
which affects performance for workloads that generate many partial
reads. This patch adds an asynchronous partial read path as well as
the required partial read ctx.
Signed-off-by: Heiner Litz
---
drivers/lightnvm/pblk-read.c
Hi Igor,
thanks for testing. You are correct with goto fail_pages
I will fix, rebase on top of 4-19 and resend the patch.
Heiner
On Wed, Jun 13, 2018 at 10:49 AM Igor Konopko wrote:
>
>
>
> On 12.06.2018 10:09, Matias Bjørling wrote:
> > On 06/12/2018 04:59 PM, Javier Gonzalez wrote:
> >>> On
On 12.06.2018 10:09, Matias Bjørling wrote:
On 06/12/2018 04:59 PM, Javier Gonzalez wrote:
On 11 Jun 2018, at 22.53, Heiner Litz wrote:
In the read path, partial reads are currently performed synchronously
which affects performance for workloads that generate many partial
reads. This patch
On 06/12/2018 04:59 PM, Javier Gonzalez wrote:
On 11 Jun 2018, at 22.53, Heiner Litz wrote:
In the read path, partial reads are currently performed synchronously
which affects performance for workloads that generate many partial
reads. This patch adds an asynchronous partial read path as well
> On 11 Jun 2018, at 22.53, Heiner Litz wrote:
>
> In the read path, partial reads are currently performed synchronously
> which affects performance for workloads that generate many partial
> reads. This patch adds an asynchronous partial read path as well as
> the required partial read ctx.
>
In the read path, partial reads are currently performed synchronously
which affects performance for workloads that generate many partial
reads. This patch adds an asynchronous partial read path as well as
the required partial read ctx.
Signed-off-by: Heiner Litz
---
drivers/lightnvm/pblk-read.c
12 matches
Mail list logo