[dpdk-dev] [EXT] Re: [PATCH v5 20/29] node: ipv4 lookup for arm64

[Nithin Dabilpuram]

On Tue, May 12, 2020 at 11:31:26AM +0200, David Marchand wrote:
> External Email
> 
> ----------------------------------------------------------------------
> On Sat, Apr 11, 2020 at 4:16 PM <jerinj at marvell.com> wrote:
> >
> > From: Nithin Dabilpuram <ndabilpuram at marvell.com>
> >
> > Add arm64 specific IPv4 lookup process function
> > for ip4_lookup node. This node performs LPM lookup
> > on every packet received and forwards it to a next
> > node that is identified by lookup result.
> >
> > Signed-off-by: Nithin Dabilpuram <ndabilpuram at marvell.com>
> > Signed-off-by: Kiran Kumar K <kirankumark at marvell.com>
> > Signed-off-by: Pavan Nikhilesh <pbhagavatula at marvell.com>
> > ---
> >  lib/librte_node/ip4_lookup.c      |   6 +
> >  lib/librte_node/ip4_lookup_neon.h | 238 ++++++++++++++++++++++++++++++
> >  2 files changed, 244 insertions(+)
> >  create mode 100644 lib/librte_node/ip4_lookup_neon.h
> 
> Checking OVS dpdk-latest branch, I caught a build issue on Ubuntu
> 16.04 for aarch64.
> I reproduced it in travis by forcing the distribution to xenial in .travis.yml.
> 
> FAILED: gcc -Ilib/lib@@rte_node at sta -Ilib -I../lib -Ilib/librte_node
> -I../lib/librte_node -I. -I../ -Iconfig -I../config
> -Ilib/librte_eal/include -I../lib/librte_eal/include
> -Ilib/librte_eal/linux/include -I../lib/librte_eal/linux/include
> -Ilib/librte_eal/arm/include -I../lib/librte_eal/arm/include
> -Ilib/librte_eal/common -I../lib/librte_eal/common -Ilib/librte_eal
> -I../lib/librte_eal -Ilib/librte_kvargs -I../lib/librte_kvargs
> -Ilib/librte_telemetry/../librte_metrics
> -I../lib/librte_telemetry/../librte_metrics -Ilib/librte_telemetry
> -I../lib/librte_telemetry -Ilib/librte_graph -I../lib/librte_graph
> -Ilib/librte_mbuf -I../lib/librte_mbuf -Ilib/librte_mempool
> -I../lib/librte_mempool -Ilib/librte_ring -I../lib/librte_ring
> -Ilib/librte_lpm -I../lib/librte_lpm -Ilib/librte_hash
> -I../lib/librte_hash -Ilib/librte_ethdev -I../lib/librte_ethdev
> -Ilib/librte_net -I../lib/librte_net -Ilib/librte_meter
> -I../lib/librte_meter -Ilib/librte_cryptodev -I../lib/librte_cryptodev
> -fdiagnostics-color=always -pipe -D_FILE_OFFSET_BITS=64 -Wall
> -Winvalid-pch -Werror -O2 -g -include rte_config.h -Wextra -Wcast-qual
> -Wdeprecated -Wformat-nonliteral -Wformat-security
> -Wmissing-declarations -Wmissing-prototypes -Wnested-externs
> -Wold-style-definition -Wpointer-arith -Wsign-compare
> -Wstrict-prototypes -Wundef -Wwrite-strings
> -Wno-missing-field-initializers -D_GNU_SOURCE -fPIC
> -DALLOW_EXPERIMENTAL_API -DALLOW_INTERNAL_API -fno-strict-aliasing
> -MD -MQ 'lib/lib@@rte_node at sta/librte_node_ip4_lookup.c.o' -MF
> 'lib/lib@@rte_node at sta/librte_node_ip4_lookup.c.o.d' -o
> 'lib/lib@@rte_node at sta/librte_node_ip4_lookup.c.o' -c
> ../lib/librte_node/ip4_lookup.c
> In file included from ../lib/librte_node/ip4_lookup.c:34:0:
> ../lib/librte_node/ip4_lookup_neon.h: In function ‘ip4_lookup_node_process’:
> ../lib/librte_node/ip4_lookup_neon.h:25:12: error: ‘dip’ may be used
> uninitialized in this function [-Werror=maybe-uninitialized]
>   int32x4_t dip;
>             ^
> cc1: all warnings being treated as errors
> ninja: build stopped: subcommand failed.
> 
> 
> The odd thing is that a more recent gcc does not complain.
> So there must be a catch, can you have a look?
> Thanks.
> 

Sure. Thanks. Will test and get back.
> 
> -- 
> David Marchand
>