Thread (2 messages) 2 messages, 2 authors, 2019-02-28

Re: [BUG] Linux-next fails to link on ARM in allmodconfig and allyesconfig

From: Leslie Monis <hidden>
Date: 2019-02-28 07:38:31

On Thu, Feb 28, 2019 at 10:29:42AM +0300, Maxim Kuvyrkov wrote:
Hi Mohit,

It seems that your commit 3f7ae5f3dc5295ac17d6521130ed8a8f8a723fbf causes a link failure for ARM in allmodconfig and allyesconfig.  This happens for at least gcc-8 compiler.
===
ERROR: "__aeabi_uldivmod" [net/sched/sch_pie.ko] undefined!
make[1]: *** [__modpost] Error 1
make: *** [modules] Error 2
===

Would you please investigate?  Below are diagnostics from Linaro's automated bisection job.

===
Culprit:
<cut>
commit 3f7ae5f3dc5295ac17d6521130ed8a8f8a723fbf
Author: Mohit P. Tahiliani [off-list ref]
Date:   Tue Feb 26 00:39:59 2019 +0530

   net: sched: pie: add more cases to auto-tune alpha and beta

   The current implementation scales the local alpha and beta
   variables in the calculate_probability function by the same
   amount for all values of drop probability below 1%.

   RFC 8033 suggests using additional cases for auto-tuning
   alpha and beta when the drop probability is less than 1%.

   In order to add more auto-tuning cases, MAX_PROB must be
   scaled by u64 instead of u32 to prevent underflow when
   scaling the local alpha and beta variables in the
   calculate_probability function.

   Signed-off-by: Mohit P. Tahiliani [off-list ref]
   Signed-off-by: Dhaval Khandla [off-list ref]
   Signed-off-by: Hrishikesh Hiraskar [off-list ref]
   Signed-off-by: Manish Kumar B [off-list ref]
   Signed-off-by: Sachin D. Patil [off-list ref]
   Signed-off-by: Leslie Monis [off-list ref]
   Acked-by: Dave Taht [off-list ref]
   Acked-by: Jamal Hadi Salim [off-list ref]
   Signed-off-by: David S. Miller [off-list ref]
</cut>

Results regressed from (for last_good == 30a92ad703b93a96588e05b5bcd7247d7350c673)
reset_artifacts:
-10
build_abe binutils:
-9
build_abe stage1:
-8
linux_n_obj:
26184
linux build successful:
all

to (for first_bad == 3f7ae5f3dc5295ac17d6521130ed8a8f8a723fbf)
reset_artifacts:
-10
build_abe binutils:
-9
build_abe stage1:
-8
linux_n_obj:
19184

First few errors in logs of first_bad:
ERROR: "__aeabi_uldivmod" [net/sched/sch_pie.ko] undefined!
make[1]: *** [__modpost] Error 1
make: *** [modules] Error 2

Artifacts of last_good build: https://ci.linaro.org/job/tcwg_kernel-bisect-gnu-release-arm-next-allmodconfig/18/artifact/artifacts/build-30a92ad703b93a96588e05b5bcd7247d7350c673/
Artifacts of first_bad build: https://ci.linaro.org/job/tcwg_kernel-bisect-gnu-release-arm-next-allmodconfig/18/artifact/artifacts/build-3f7ae5f3dc5295ac17d6521130ed8a8f8a723fbf/
===

--
Maxim Kuvyrkov
www.linaro.org

Hi Max,

This bug was fixed by commit ff8285f81822dc8f528b36b6c5c8ab132367e92d

commit ff8285f81822dc8f528b36b6c5c8ab132367e92d
Author: Leslie Monis [off-list ref]
Date:   Wed Feb 27 06:30:06 2019 +0530

    net: sched: pie: fix 64-bit division
    
    Use div_u64() to resolve build failures on 32-bit platforms.
    
    Fixes: 3f7ae5f3dc52 ("net: sched: pie: add more cases to auto-tune alpha and beta")
    Signed-off-by: Leslie Monis [off-list ref]
    Reported-by: Randy Dunlap [off-list ref]
    Tested-by: Randy Dunlap [off-list ref]
    Signed-off-by: David S. Miller [off-list ref]

Leslie

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help