projects / thirdparty / kernel / stable-queue.git / blame
| Commit | Line | Data |
|---|---|---|
| 91562c7b GKH |
1 | From foo@baz Wed 22 May 2019 08:37:51 AM CEST |
| 2 | From: Florian Fainelli <f.fainelli@gmail.com> | |
| 3 | Date: Mon, 13 May 2019 14:06:24 -0700 | |
| 4 | Subject: net: Always descend into dsa/ | |
| 5 | ||
| 6 | From: Florian Fainelli <f.fainelli@gmail.com> | |
| 7 | ||
| 8 | [ Upstream commit 0fe9f173d6cda95874edeb413b1fa9907b5ae830 ] | |
| 9 | ||
| 10 | Jiri reported that with a kernel built with CONFIG_FIXED_PHY=y, | |
| 11 | CONFIG_NET_DSA=m and CONFIG_NET_DSA_LOOP=m, we would not get to a | |
| 12 | functional state where the mock-up driver is registered. Turns out that | |
| 13 | we are not descending into drivers/net/dsa/ unconditionally, and we | |
| 14 | won't be able to link-in dsa_loop_bdinfo.o which does the actual mock-up | |
| 15 | mdio device registration. | |
| 16 | ||
| 17 | Reported-by: Jiri Pirko <jiri@resnulli.us> | |
| 18 | Fixes: 40013ff20b1b ("net: dsa: Fix functional dsa-loop dependency on FIXED_PHY") | |
| 19 | Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> | |
| 20 | Reviewed-by: Vivien Didelot <vivien.didelot@gmail.com> | |
| 21 | Tested-by: Jiri Pirko <jiri@resnulli.us> | |
| 22 | Signed-off-by: David S. Miller <davem@davemloft.net> | |
| 23 | Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> | |
| 24 | --- | |
| 25 | drivers/net/Makefile | 2 +- | |
| 26 | 1 file changed, 1 insertion(+), 1 deletion(-) | |
| 27 | ||
| 28 | --- a/drivers/net/Makefile | |
| 29 | +++ b/drivers/net/Makefile | |
| 30 | @@ -40,7 +40,7 @@ obj-$(CONFIG_ARCNET) += arcnet/ | |
| 31 | obj-$(CONFIG_DEV_APPLETALK) += appletalk/ | |
| 32 | obj-$(CONFIG_CAIF) += caif/ | |
| 33 | obj-$(CONFIG_CAN) += can/ | |
| 34 | -obj-$(CONFIG_NET_DSA) += dsa/ | |
| 35 | +obj-y += dsa/ | |
| 36 | obj-$(CONFIG_ETHERNET) += ethernet/ | |
| 37 | obj-$(CONFIG_FDDI) += fddi/ | |
| 38 | obj-$(CONFIG_HIPPI) += hippi/ |