stable-rc 4.9.183-rc1/582089d9e746: regressions detected in project stable v4.9.y on OE - sanity
Summary ------------------------------------------------------------------------
kernel: 4.9.183-rc1 git repo: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git git branch: linux-4.9.y git commit: 582089d9e746308b2feadf7b7f72cfd40a241c8f git describe: v4.9.182-118-g582089d9e746 Test details: https://qa-reports.linaro.org/lkft/linux-stable-rc-4.9-oe-sanity/build/v4.9....
Regressions (compared to build v4.9.182-90-g021209965fa2) ------------------------------------------------------------------------
dragonboard-410c: build: * build_process
hi6220-hikey: build: * build_process
juno-r2 - arm64: build: * build_process
x15 - arm: build: * build_process
No fixes (compared to build v4.9.182-90-g021209965fa2)
Boards, architectures and test suites: -------------------------------------
dragonboard-410c * build - fail: 1,
hi6220-hikey * build - fail: 1,
i386 * boot-lkft-ltp-quickhit-linux-4.9.y-594 - pass: 1 * build - pass: 1 * ltp-quickhit-tests - pass: 109
juno-r2 - arm64 * build - fail: 1,
x15 - arm * build - fail: 1,
x86_64 * boot-lkft-ltp-quickhit-linux-4.9.y-594 - pass: 1 * build - pass: 1 * ltp-quickhit-tests - skip: 1, pass: 108
Hello!
On Thu, 20 Jun 2019 at 10:37, Linaro QA qa-reports@linaro.org wrote:
Summary
kernel: 4.9.183-rc1 git repo: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git git branch: linux-4.9.y git commit: 582089d9e746308b2feadf7b7f72cfd40a241c8f git describe: v4.9.182-118-g582089d9e746 Test details: https://qa-reports.linaro.org/lkft/linux-stable-rc-4.9-oe-sanity/build/v4.9....
Regressions (compared to build v4.9.182-90-g021209965fa2)
dragonboard-410c: build: * build_process
hi6220-hikey: build: * build_process
juno-r2 - arm64: build: * build_process
x15 - arm: build: * build_process
All Arm builds (32/64 bits) failed due to: | /linux/mm/khugepaged.c: In function 'collapse_huge_page': | /linux/mm/khugepaged.c:1008:7: error: implicit declaration of function 'mmget_still_valid' [-Werror=implicit-function-declaration]
That function (mmget_still_valid) comes from 04f5866e41fb70690e28397487d8bd8eea7d712a in include/linux/sched/mm.h: -----8<----------8<----------8<----- static inline bool mmget_still_valid(struct mm_struct *mm) { return likely(!mm->core_state); } ----->8---------->8---------->8-----
I verified it builds by applying this patch: -----8<----------8<----------8<-----diff --git a/mm/khugepaged.c b/mm/khugepaged.c index 8217ee5d66ef..eef5a67193c2 100644 --- a/mm/khugepaged.c +++ b/mm/khugepaged.c @@ -1005,7 +1005,7 @@ static void collapse_huge_page(struct mm_struct *mm, */ down_write(&mm->mmap_sem); result = SCAN_ANY_PROCESS; - if (!mmget_still_valid(mm)) + if (!likely(!mm->core_state)) goto out; result = hugepage_vma_revalidate(mm, address, &vma); if (result) ----->8---------->8---------->8-----
Greetings!
Daniel Díaz daniel.diaz@linaro.org
No fixes (compared to build v4.9.182-90-g021209965fa2)
Boards, architectures and test suites:
dragonboard-410c
- build - fail: 1,
hi6220-hikey
- build - fail: 1,
i386
- boot-lkft-ltp-quickhit-linux-4.9.y-594 - pass: 1
- build - pass: 1
- ltp-quickhit-tests - pass: 109
juno-r2 - arm64
- build - fail: 1,
x15 - arm
- build - fail: 1,
x86_64
- boot-lkft-ltp-quickhit-linux-4.9.y-594 - pass: 1
- build - pass: 1
- ltp-quickhit-tests - skip: 1, pass: 108
-- Linaro QA (BETA) https://qa-reports.linaro.org
On Thu, Jun 20, 2019 at 10:47:15AM -0500, Daniel Díaz wrote:
Hello!
On Thu, 20 Jun 2019 at 10:37, Linaro QA qa-reports@linaro.org wrote:
Summary
kernel: 4.9.183-rc1 git repo: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git git branch: linux-4.9.y git commit: 582089d9e746308b2feadf7b7f72cfd40a241c8f git describe: v4.9.182-118-g582089d9e746 Test details: https://qa-reports.linaro.org/lkft/linux-stable-rc-4.9-oe-sanity/build/v4.9....
Regressions (compared to build v4.9.182-90-g021209965fa2)
dragonboard-410c: build: * build_process
hi6220-hikey: build: * build_process
juno-r2 - arm64: build: * build_process
x15 - arm: build: * build_process
All Arm builds (32/64 bits) failed due to: | /linux/mm/khugepaged.c: In function 'collapse_huge_page': | /linux/mm/khugepaged.c:1008:7: error: implicit declaration of function 'mmget_still_valid' [-Werror=implicit-function-declaration]
That function (mmget_still_valid) comes from 04f5866e41fb70690e28397487d8bd8eea7d712a in include/linux/sched/mm.h: -----8<----------8<----------8<----- static inline bool mmget_still_valid(struct mm_struct *mm) { return likely(!mm->core_state); } ----->8---------->8---------->8-----
I verified it builds by applying this patch: -----8<----------8<----------8<-----diff --git a/mm/khugepaged.c b/mm/khugepaged.c index 8217ee5d66ef..eef5a67193c2 100644 --- a/mm/khugepaged.c +++ b/mm/khugepaged.c @@ -1005,7 +1005,7 @@ static void collapse_huge_page(struct mm_struct *mm, */ down_write(&mm->mmap_sem); result = SCAN_ANY_PROCESS;
if (!mmget_still_valid(mm))
if (!likely(!mm->core_state)) goto out; result = hugepage_vma_revalidate(mm, address, &vma); if (result)----->8---------->8---------->8-----
That's the correct backport? Ah, yeah, maybe. Anyway, I just dropped this patch, sorry about that, and asked for a "real" backport, so that the mm developers agree with this.
thanks,
greg k-h
-
Daniel Díaz -
Greg Kroah-Hartman -
Linaro QA