Hi Cristian,
Responses in-line below...
> -Original Message-
> From: Dumitrescu, Cristian [mailto:cristian.dumitre...@intel.com]
> Sent: Tuesday, January 02, 2018 4:14 PM
> To: alangordonde...@gmail.com
> Cc: dev@dpdk.org; Alan Dewar
> Subject: RE: [PATCH v2] sched: fix overflow errors
Hi Alan,
NAK for now.
There is a good reason for truncating the WRR cost to 8-bit value, which is
keeping the size of the rte_sched_pipe structure to single cache line (64
bytes). This is done for performance reasons at the expense of some accuracy
loss for the scheduling of the 4x queues per
On Thu, 2017-11-30 at 09:04 +, alangordonde...@gmail.com wrote:
> From: Alan Dewar
>
> Revised patch - this version fixes an issue when a small wrr_cost is
> shifted so far right that its value becomes zero.
>
> The WRR code calculates the lowest common denominator between the
> four
> WRR w
3 matches
Mail list logo