On 4/11/20 5:13 PM, David Bauer wrote: > Hello Paul, > > On 4/3/20 10:58 PM, Paul Blazejowski wrote: >> Hello David, >> >> It is great to hear to we're switching to 5.4 on ath79 target but i >> wanted to bring to your attention that at least on the NAND subtargets >> (for me wndr3700v4) we need this commit [1] from Kevin to get the router >> to boot otherwise there's an oops in do dsp which prevents my router to >> boot and produces bootloop, Kevin's patch is incomplete as i have also >> alter it locally to apply to 5.4 kernel. With Michal's mtd concat patch >> i have successfully run the kernel/build for a couple of days with no >> issues. Maybe Michal can chaim in to confirm as he uses the same >> hardware in WNDR4300 routers just to be sure? > > I've bisected this issue a bit further. Changing the CPU type for the > whole target doesn't seem right, as boards prior to AR934x are 24kc. > > I've isolated the issue we are facing to 308-mips32r2_tune.patch. I'm a > bit puzzled how to proceed as i suspect other targets using a 74kc arch > will most likely show the same issue. > > I propose to revert the change for ath79 locally using a target patch to > allow the target to be upgraded to Kernel 5.4. > > IMHO 308-mips32r2_tune.patch should be a target specific patch anyhow, > as optimizing for a single CPU arch for all affected targets might not > be very efficient. But I'm not all to familiar with this topic. > > Best wishes > David
Hi, I also get this problem with mainline kernel. See here for some more details: https://bugs.openwrt.org/index.php?do=details&task_id=2928 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94506 Hauke _______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/mailman/listinfo/openwrt-devel
