- Linux-stable-mirror - lists.linaro.org

Patch "mm/vmscan: wake up flushers for legacy cgroups too" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled mm/vmscan: wake up flushers for legacy cgroups too to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: mm-vmscan-wake-up-flushers-for-legacy-cgroups-too.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 1c610d5f93c709df56787f50b3576704ac271826 Mon Sep 17 00:00:00 2001 From: Andrey Ryabinin <aryabinin(a)virtuozzo.com> Date: Thu, 22 Mar 2018 16:17:42 -0700 Subject: mm/vmscan: wake up flushers for legacy cgroups too From: Andrey Ryabinin <aryabinin(a)virtuozzo.com> commit 1c610d5f93c709df56787f50b3576704ac271826 upstream. Commit 726d061fbd36 ("mm: vmscan: kick flushers when we encounter dirty pages on the LRU") added flusher invocation to shrink_inactive_list() when many dirty pages on the LRU are encountered. However, shrink_inactive_list() doesn't wake up flushers for legacy cgroup reclaim, so the next commit bbef938429f5 ("mm: vmscan: remove old flusher wakeup from direct reclaim path") removed the only source of flusher's wake up in legacy mem cgroup reclaim path. This leads to premature OOM if there is too many dirty pages in cgroup: # mkdir /sys/fs/cgroup/memory/test # echo $$ > /sys/fs/cgroup/memory/test/tasks # echo 50M > /sys/fs/cgroup/memory/test/memory.limit_in_bytes # dd if=/dev/zero of=tmp_file bs=1M count=100 Killed dd invoked oom-killer: gfp_mask=0x14000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=0 Call Trace: dump_stack+0x46/0x65 dump_header+0x6b/0x2ac oom_kill_process+0x21c/0x4a0 out_of_memory+0x2a5/0x4b0 mem_cgroup_out_of_memory+0x3b/0x60 mem_cgroup_oom_synchronize+0x2ed/0x330 pagefault_out_of_memory+0x24/0x54 __do_page_fault+0x521/0x540 page_fault+0x45/0x50 Task in /test killed as a result of limit of /test memory: usage 51200kB, limit 51200kB, failcnt 73 memory+swap: usage 51200kB, limit 9007199254740988kB, failcnt 0 kmem: usage 296kB, limit 9007199254740988kB, failcnt 0 Memory cgroup stats for /test: cache:49632KB rss:1056KB rss_huge:0KB shmem:0KB mapped_file:0KB dirty:49500KB writeback:0KB swap:0KB inactive_anon:0KB active_anon:1168KB inactive_file:24760KB active_file:24960KB unevictable:0KB Memory cgroup out of memory: Kill process 3861 (bash) score 88 or sacrifice child Killed process 3876 (dd) total-vm:8484kB, anon-rss:1052kB, file-rss:1720kB, shmem-rss:0kB oom_reaper: reaped process 3876 (dd), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB Wake up flushers in legacy cgroup reclaim too. Link: http://lkml.kernel.org/r/20180315164553.17856-1-aryabinin@virtuozzo.com Fixes: bbef938429f5 ("mm: vmscan: remove old flusher wakeup from direct reclaim path") Signed-off-by: Andrey Ryabinin <aryabinin(a)virtuozzo.com> Tested-by: Shakeel Butt <shakeelb(a)google.com> Acked-by: Michal Hocko <mhocko(a)suse.cz> Cc: Mel Gorman <mgorman(a)techsingularity.net> Cc: Tejun Heo <tj(a)kernel.org> Cc: Johannes Weiner <hannes(a)cmpxchg.org> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- mm/vmscan.c | 31 ++++++++++++++++--------------- 1 file changed, 16 insertions(+), 15 deletions(-) --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -1847,6 +1847,20 @@ shrink_inactive_list(unsigned long nr_to set_bit(PGDAT_WRITEBACK, &pgdat->flags); /* + * If dirty pages are scanned that are not queued for IO, it + * implies that flushers are not doing their job. This can + * happen when memory pressure pushes dirty pages to the end of + * the LRU before the dirty limits are breached and the dirty + * data has expired. It can also happen when the proportion of + * dirty pages grows not through writes but through memory + * pressure reclaiming all the clean cache. And in some cases, + * the flushers simply cannot keep up with the allocation + * rate. Nudge the flusher threads in case they are asleep. + */ + if (stat.nr_unqueued_dirty == nr_taken) + wakeup_flusher_threads(WB_REASON_VMSCAN); + + /* * Legacy memcg will stall in page writeback so avoid forcibly * stalling here. */ @@ -1858,22 +1872,9 @@ shrink_inactive_list(unsigned long nr_to if (stat.nr_dirty && stat.nr_dirty == stat.nr_congested) set_bit(PGDAT_CONGESTED, &pgdat->flags); - /* - * If dirty pages are scanned that are not queued for IO, it - * implies that flushers are not doing their job. This can - * happen when memory pressure pushes dirty pages to the end of - * the LRU before the dirty limits are breached and the dirty - * data has expired. It can also happen when the proportion of - * dirty pages grows not through writes but through memory - * pressure reclaiming all the clean cache. And in some cases, - * the flushers simply cannot keep up with the allocation - * rate. Nudge the flusher threads in case they are asleep, but - * also allow kswapd to start writing pages during reclaim. - */ - if (stat.nr_unqueued_dirty == nr_taken) { - wakeup_flusher_threads(WB_REASON_VMSCAN); + /* Allow kswapd to start writing pages during reclaim. */ + if (stat.nr_unqueued_dirty == nr_taken) set_bit(PGDAT_DIRTY, &pgdat->flags); - } /* * If kswapd scans pages marked marked for immediate Patches currently in stable-queue which might be from aryabinin(a)virtuozzo.com are queue-4.15/mm-vmscan-wake-up-flushers-for-legacy-cgroups-too.patch

7 years, 9 months

1
0
0 0

Patch "mm/vmalloc: add interfaces to free unmapped page table" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled mm/vmalloc: add interfaces to free unmapped page table to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: mm-vmalloc-add-interfaces-to-free-unmapped-page-table.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From b6bdb7517c3d3f41f20e5c2948d6bc3f8897394e Mon Sep 17 00:00:00 2001 From: Toshi Kani <toshi.kani(a)hpe.com> Date: Thu, 22 Mar 2018 16:17:20 -0700 Subject: mm/vmalloc: add interfaces to free unmapped page table From: Toshi Kani <toshi.kani(a)hpe.com> commit b6bdb7517c3d3f41f20e5c2948d6bc3f8897394e upstream. On architectures with CONFIG_HAVE_ARCH_HUGE_VMAP set, ioremap() may create pud/pmd mappings. A kernel panic was observed on arm64 systems with Cortex-A75 in the following steps as described by Hanjun Guo. 1. ioremap a 4K size, valid page table will build, 2. iounmap it, pte0 will set to 0; 3. ioremap the same address with 2M size, pgd/pmd is unchanged, then set the a new value for pmd; 4. pte0 is leaked; 5. CPU may meet exception because the old pmd is still in TLB, which will lead to kernel panic. This panic is not reproducible on x86. INVLPG, called from iounmap, purges all levels of entries associated with purged address on x86. x86 still has memory leak. The patch changes the ioremap path to free unmapped page table(s) since doing so in the unmap path has the following issues: - The iounmap() path is shared with vunmap(). Since vmap() only supports pte mappings, making vunmap() to free a pte page is an overhead for regular vmap users as they do not need a pte page freed up. - Checking if all entries in a pte page are cleared in the unmap path is racy, and serializing this check is expensive. - The unmap path calls free_vmap_area_noflush() to do lazy TLB purges. Clearing a pud/pmd entry before the lazy TLB purges needs extra TLB purge. Add two interfaces, pud_free_pmd_page() and pmd_free_pte_page(), which clear a given pud/pmd entry and free up a page for the lower level entries. This patch implements their stub functions on x86 and arm64, which work as workaround. [akpm(a)linux-foundation.org: fix typo in pmd_free_pte_page() stub] Link: http://lkml.kernel.org/r/20180314180155.19492-2-toshi.kani@hpe.com Fixes: e61ce6ade404e ("mm: change ioremap to set up huge I/O mappings") Reported-by: Lei Li <lious.lilei(a)hisilicon.com> Signed-off-by: Toshi Kani <toshi.kani(a)hpe.com> Cc: Catalin Marinas <catalin.marinas(a)arm.com> Cc: Wang Xuefeng <wxf.wang(a)hisilicon.com> Cc: Will Deacon <will.deacon(a)arm.com> Cc: Hanjun Guo <guohanjun(a)huawei.com> Cc: Michal Hocko <mhocko(a)suse.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Cc: Ingo Molnar <mingo(a)redhat.com> Cc: "H. Peter Anvin" <hpa(a)zytor.com> Cc: Borislav Petkov <bp(a)suse.de> Cc: Matthew Wilcox <willy(a)infradead.org> Cc: Chintan Pandya <cpandya(a)codeaurora.org> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/arm64/mm/mmu.c | 10 ++++++++++ arch/x86/mm/pgtable.c | 24 ++++++++++++++++++++++++ include/asm-generic/pgtable.h | 10 ++++++++++ lib/ioremap.c | 6 ++++-- 4 files changed, 48 insertions(+), 2 deletions(-) --- a/arch/arm64/mm/mmu.c +++ b/arch/arm64/mm/mmu.c @@ -941,3 +941,13 @@ int pmd_clear_huge(pmd_t *pmd) pmd_clear(pmd); return 1; } + +int pud_free_pmd_page(pud_t *pud) +{ + return pud_none(*pud); +} + +int pmd_free_pte_page(pmd_t *pmd) +{ + return pmd_none(*pmd); +} --- a/arch/x86/mm/pgtable.c +++ b/arch/x86/mm/pgtable.c @@ -702,4 +702,28 @@ int pmd_clear_huge(pmd_t *pmd) return 0; } + +/** + * pud_free_pmd_page - Clear pud entry and free pmd page. + * @pud: Pointer to a PUD. + * + * Context: The pud range has been unmaped and TLB purged. + * Return: 1 if clearing the entry succeeded. 0 otherwise. + */ +int pud_free_pmd_page(pud_t *pud) +{ + return pud_none(*pud); +} + +/** + * pmd_free_pte_page - Clear pmd entry and free pte page. + * @pmd: Pointer to a PMD. + * + * Context: The pmd range has been unmaped and TLB purged. + * Return: 1 if clearing the entry succeeded. 0 otherwise. + */ +int pmd_free_pte_page(pmd_t *pmd) +{ + return pmd_none(*pmd); +} #endif /* CONFIG_HAVE_ARCH_HUGE_VMAP */ --- a/include/asm-generic/pgtable.h +++ b/include/asm-generic/pgtable.h @@ -976,6 +976,8 @@ int pud_set_huge(pud_t *pud, phys_addr_t int pmd_set_huge(pmd_t *pmd, phys_addr_t addr, pgprot_t prot); int pud_clear_huge(pud_t *pud); int pmd_clear_huge(pmd_t *pmd); +int pud_free_pmd_page(pud_t *pud); +int pmd_free_pte_page(pmd_t *pmd); #else /* !CONFIG_HAVE_ARCH_HUGE_VMAP */ static inline int p4d_set_huge(p4d_t *p4d, phys_addr_t addr, pgprot_t prot) { @@ -1001,6 +1003,14 @@ static inline int pmd_clear_huge(pmd_t * { return 0; } +static inline int pud_free_pmd_page(pud_t *pud) +{ + return 0; +} +static inline int pmd_free_pte_page(pmd_t *pmd) +{ + return 0; +} #endif /* CONFIG_HAVE_ARCH_HUGE_VMAP */ #ifndef __HAVE_ARCH_FLUSH_PMD_TLB_RANGE --- a/lib/ioremap.c +++ b/lib/ioremap.c @@ -91,7 +91,8 @@ static inline int ioremap_pmd_range(pud_ if (ioremap_pmd_enabled() && ((next - addr) == PMD_SIZE) && - IS_ALIGNED(phys_addr + addr, PMD_SIZE)) { + IS_ALIGNED(phys_addr + addr, PMD_SIZE) && + pmd_free_pte_page(pmd)) { if (pmd_set_huge(pmd, phys_addr + addr, prot)) continue; } @@ -117,7 +118,8 @@ static inline int ioremap_pud_range(p4d_ if (ioremap_pud_enabled() && ((next - addr) == PUD_SIZE) && - IS_ALIGNED(phys_addr + addr, PUD_SIZE)) { + IS_ALIGNED(phys_addr + addr, PUD_SIZE) && + pud_free_pmd_page(pud)) { if (pud_set_huge(pud, phys_addr + addr, prot)) continue; } Patches currently in stable-queue which might be from toshi.kani(a)hpe.com are queue-4.15/mm-vmalloc-add-interfaces-to-free-unmapped-page-table.patch queue-4.15/x86-mm-implement-free-pmd-pte-page-interfaces.patch

7 years, 9 months

1
0
0 0

Patch "mm/thp: do not wait for lock_page() in deferred_split_scan()" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled mm/thp: do not wait for lock_page() in deferred_split_scan() to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: mm-thp-do-not-wait-for-lock_page-in-deferred_split_scan.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From fa41b900c30b45fab03783724932dc30cd46a6be Mon Sep 17 00:00:00 2001 From: "Kirill A. Shutemov" <kirill.shutemov(a)linux.intel.com> Date: Thu, 22 Mar 2018 16:17:31 -0700 Subject: mm/thp: do not wait for lock_page() in deferred_split_scan() From: Kirill A. Shutemov <kirill.shutemov(a)linux.intel.com> commit fa41b900c30b45fab03783724932dc30cd46a6be upstream. deferred_split_scan() gets called from reclaim path. Waiting for page lock may lead to deadlock there. Replace lock_page() with trylock_page() and skip the page if we failed to lock it. We will get to the page on the next scan. Link: http://lkml.kernel.org/r/20180315150747.31945-1-kirill.shutemov@linux.intel… Fixes: 9a982250f773 ("thp: introduce deferred_split_huge_page()") Signed-off-by: Kirill A. Shutemov <kirill.shutemov(a)linux.intel.com> Acked-by: Michal Hocko <mhocko(a)suse.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- mm/huge_memory.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) --- a/mm/huge_memory.c +++ b/mm/huge_memory.c @@ -2791,11 +2791,13 @@ static unsigned long deferred_split_scan list_for_each_safe(pos, next, &list) { page = list_entry((void *)pos, struct page, mapping); - lock_page(page); + if (!trylock_page(page)) + goto next; /* split_huge_page() removes page from list on success */ if (!split_huge_page(page)) split++; unlock_page(page); +next: put_page(page); } Patches currently in stable-queue which might be from kirill.shutemov(a)linux.intel.com are queue-4.15/mm-khugepaged.c-convert-vm_bug_on-to-collapse-fail.patch queue-4.15/mm-thp-do-not-wait-for-lock_page-in-deferred_split_scan.patch queue-4.15/hugetlbfs-check-for-pgoff-value-overflow.patch queue-4.15/mm-shmem-do-not-wait-for-lock_page-in-shmem_unused_huge_shrink.patch

7 years, 9 months

1
0
0 0

Patch "mm/shmem: do not wait for lock_page() in shmem_unused_huge_shrink()" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled mm/shmem: do not wait for lock_page() in shmem_unused_huge_shrink() to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: mm-shmem-do-not-wait-for-lock_page-in-shmem_unused_huge_shrink.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From b3cd54b257ad95d344d121dc563d943ca39b0921 Mon Sep 17 00:00:00 2001 From: "Kirill A. Shutemov" <kirill.shutemov(a)linux.intel.com> Date: Thu, 22 Mar 2018 16:17:35 -0700 Subject: mm/shmem: do not wait for lock_page() in shmem_unused_huge_shrink() From: Kirill A. Shutemov <kirill.shutemov(a)linux.intel.com> commit b3cd54b257ad95d344d121dc563d943ca39b0921 upstream. shmem_unused_huge_shrink() gets called from reclaim path. Waiting for page lock may lead to deadlock there. There was a bug report that may be attributed to this: http://lkml.kernel.org/r/alpine.LRH.2.11.1801242349220.30642@mail.ewheeler.… Replace lock_page() with trylock_page() and skip the page if we failed to lock it. We will get to the page on the next scan. We can test for the PageTransHuge() outside the page lock as we only need protection against splitting the page under us. Holding pin oni the page is enough for this. Link: http://lkml.kernel.org/r/20180316210830.43738-1-kirill.shutemov@linux.intel… Fixes: 779750d20b93 ("shmem: split huge pages beyond i_size under memory pressure") Signed-off-by: Kirill A. Shutemov <kirill.shutemov(a)linux.intel.com> Reported-by: Eric Wheeler <linux-mm(a)lists.ewheeler.net> Acked-by: Michal Hocko <mhocko(a)suse.com> Reviewed-by: Andrew Morton <akpm(a)linux-foundation.org> Cc: Tetsuo Handa <penguin-kernel(a)I-love.SAKURA.ne.jp> Cc: Hugh Dickins <hughd(a)google.com> Cc: <stable(a)vger.kernel.org> [4.8+] Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- mm/shmem.c | 31 ++++++++++++++++++++----------- 1 file changed, 20 insertions(+), 11 deletions(-) --- a/mm/shmem.c +++ b/mm/shmem.c @@ -493,36 +493,45 @@ next: info = list_entry(pos, struct shmem_inode_info, shrinklist); inode = &info->vfs_inode; - if (nr_to_split && split >= nr_to_split) { - iput(inode); - continue; - } + if (nr_to_split && split >= nr_to_split) + goto leave; - page = find_lock_page(inode->i_mapping, + page = find_get_page(inode->i_mapping, (inode->i_size & HPAGE_PMD_MASK) >> PAGE_SHIFT); if (!page) goto drop; + /* No huge page at the end of the file: nothing to split */ if (!PageTransHuge(page)) { - unlock_page(page); put_page(page); goto drop; } + /* + * Leave the inode on the list if we failed to lock + * the page at this time. + * + * Waiting for the lock may lead to deadlock in the + * reclaim path. + */ + if (!trylock_page(page)) { + put_page(page); + goto leave; + } + ret = split_huge_page(page); unlock_page(page); put_page(page); - if (ret) { - /* split failed: leave it on the list */ - iput(inode); - continue; - } + /* If split failed leave the inode on the list */ + if (ret) + goto leave; split++; drop: list_del_init(&info->shrinklist); removed++; +leave: iput(inode); } Patches currently in stable-queue which might be from kirill.shutemov(a)linux.intel.com are queue-4.15/mm-khugepaged.c-convert-vm_bug_on-to-collapse-fail.patch queue-4.15/mm-thp-do-not-wait-for-lock_page-in-deferred_split_scan.patch queue-4.15/hugetlbfs-check-for-pgoff-value-overflow.patch queue-4.15/mm-shmem-do-not-wait-for-lock_page-in-shmem_unused_huge_shrink.patch

7 years, 9 months

1
0
0 0

Patch "mm/khugepaged.c: convert VM_BUG_ON() to collapse fail" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled mm/khugepaged.c: convert VM_BUG_ON() to collapse fail to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: mm-khugepaged.c-convert-vm_bug_on-to-collapse-fail.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From fece2029a9e65b9a990831afe2a2b83290cbbe26 Mon Sep 17 00:00:00 2001 From: "Kirill A. Shutemov" <kirill.shutemov(a)linux.intel.com> Date: Thu, 22 Mar 2018 16:17:28 -0700 Subject: mm/khugepaged.c: convert VM_BUG_ON() to collapse fail From: Kirill A. Shutemov <kirill.shutemov(a)linux.intel.com> commit fece2029a9e65b9a990831afe2a2b83290cbbe26 upstream. khugepaged is not yet able to convert PTE-mapped huge pages back to PMD mapped. We do not collapse such pages. See check khugepaged_scan_pmd(). But if between khugepaged_scan_pmd() and __collapse_huge_page_isolate() somebody managed to instantiate THP in the range and then split the PMD back to PTEs we would have a problem -- VM_BUG_ON_PAGE(PageCompound(page)) will get triggered. It's possible since we drop mmap_sem during collapse to re-take for write. Replace the VM_BUG_ON() with graceful collapse fail. Link: http://lkml.kernel.org/r/20180315152353.27989-1-kirill.shutemov@linux.intel… Fixes: b1caa957ae6d ("khugepaged: ignore pmd tables with THP mapped with ptes") Signed-off-by: Kirill A. Shutemov <kirill.shutemov(a)linux.intel.com> Cc: Laura Abbott <labbott(a)redhat.com> Cc: Jerome Marchand <jmarchan(a)redhat.com> Cc: Vlastimil Babka <vbabka(a)suse.cz> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- mm/khugepaged.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) --- a/mm/khugepaged.c +++ b/mm/khugepaged.c @@ -530,7 +530,12 @@ static int __collapse_huge_page_isolate( goto out; } - VM_BUG_ON_PAGE(PageCompound(page), page); + /* TODO: teach khugepaged to collapse THP mapped with pte */ + if (PageCompound(page)) { + result = SCAN_PAGE_COMPOUND; + goto out; + } + VM_BUG_ON_PAGE(!PageAnon(page), page); /* Patches currently in stable-queue which might be from kirill.shutemov(a)linux.intel.com are queue-4.15/mm-khugepaged.c-convert-vm_bug_on-to-collapse-fail.patch queue-4.15/mm-thp-do-not-wait-for-lock_page-in-deferred_split_scan.patch queue-4.15/hugetlbfs-check-for-pgoff-value-overflow.patch queue-4.15/mm-shmem-do-not-wait-for-lock_page-in-shmem_unused_huge_shrink.patch

7 years, 9 months

1
0
0 0

Patch "libnvdimm, {btt, blk}: do integrity setup before add_disk()" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled libnvdimm, {btt, blk}: do integrity setup before add_disk() to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: libnvdimm-btt-blk-do-integrity-setup-before-add_disk.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 3ffb0ba9b567a8efb9a04ed3d1ec15ff333ada22 Mon Sep 17 00:00:00 2001 From: Vishal Verma <vishal.l.verma(a)intel.com> Date: Mon, 5 Mar 2018 16:56:13 -0700 Subject: libnvdimm, {btt, blk}: do integrity setup before add_disk() From: Vishal Verma <vishal.l.verma(a)intel.com> commit 3ffb0ba9b567a8efb9a04ed3d1ec15ff333ada22 upstream. Prior to 25520d55cdb6 ("block: Inline blk_integrity in struct gendisk") we needed to temporarily add a zero-capacity disk before registering for blk-integrity. But adding a zero-capacity disk caused the partition table scanning to bail early, and this resulted in partitions not coming up after a probe of the BTT or blk namespaces. We can now register for integrity before the disk has been added, and this fixes the rescan problems. Fixes: 25520d55cdb6 ("block: Inline blk_integrity in struct gendisk") Reported-by: Dariusz Dokupil <dariusz.dokupil(a)intel.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Vishal Verma <vishal.l.verma(a)intel.com> Signed-off-by: Dan Williams <dan.j.williams(a)intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/nvdimm/blk.c | 3 +-- drivers/nvdimm/btt.c | 3 +-- 2 files changed, 2 insertions(+), 4 deletions(-) --- a/drivers/nvdimm/blk.c +++ b/drivers/nvdimm/blk.c @@ -278,8 +278,6 @@ static int nsblk_attach_disk(struct nd_n disk->queue = q; disk->flags = GENHD_FL_EXT_DEVT; nvdimm_namespace_disk_name(&nsblk->common, disk->disk_name); - set_capacity(disk, 0); - device_add_disk(dev, disk); if (devm_add_action_or_reset(dev, nd_blk_release_disk, disk)) return -ENOMEM; @@ -292,6 +290,7 @@ static int nsblk_attach_disk(struct nd_n } set_capacity(disk, available_disk_size >> SECTOR_SHIFT); + device_add_disk(dev, disk); revalidate_disk(disk); return 0; } --- a/drivers/nvdimm/btt.c +++ b/drivers/nvdimm/btt.c @@ -1545,8 +1545,6 @@ static int btt_blk_init(struct btt *btt) queue_flag_set_unlocked(QUEUE_FLAG_NONROT, btt->btt_queue); btt->btt_queue->queuedata = btt; - set_capacity(btt->btt_disk, 0); - device_add_disk(&btt->nd_btt->dev, btt->btt_disk); if (btt_meta_size(btt)) { int rc = nd_integrity_init(btt->btt_disk, btt_meta_size(btt)); @@ -1558,6 +1556,7 @@ static int btt_blk_init(struct btt *btt) } } set_capacity(btt->btt_disk, btt->nlba * btt->sector_size >> 9); + device_add_disk(&btt->nd_btt->dev, btt->btt_disk); btt->nd_btt->size = btt->nlba * (u64)btt->sector_size; revalidate_disk(btt->btt_disk); Patches currently in stable-queue which might be from vishal.l.verma(a)intel.com are queue-4.15/libnvdimm-btt-blk-do-integrity-setup-before-add_disk.patch

7 years, 9 months

1
0
0 0

Patch "hugetlbfs: check for pgoff value overflow" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled hugetlbfs: check for pgoff value overflow to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: hugetlbfs-check-for-pgoff-value-overflow.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 63489f8e821144000e0bdca7e65a8d1cc23a7ee7 Mon Sep 17 00:00:00 2001 From: Mike Kravetz <mike.kravetz(a)oracle.com> Date: Thu, 22 Mar 2018 16:17:13 -0700 Subject: hugetlbfs: check for pgoff value overflow From: Mike Kravetz <mike.kravetz(a)oracle.com> commit 63489f8e821144000e0bdca7e65a8d1cc23a7ee7 upstream. A vma with vm_pgoff large enough to overflow a loff_t type when converted to a byte offset can be passed via the remap_file_pages system call. The hugetlbfs mmap routine uses the byte offset to calculate reservations and file size. A sequence such as: mmap(0x20a00000, 0x600000, 0, 0x66033, -1, 0); remap_file_pages(0x20a00000, 0x600000, 0, 0x20000000000000, 0); will result in the following when task exits/file closed, kernel BUG at mm/hugetlb.c:749! Call Trace: hugetlbfs_evict_inode+0x2f/0x40 evict+0xcb/0x190 __dentry_kill+0xcb/0x150 __fput+0x164/0x1e0 task_work_run+0x84/0xa0 exit_to_usermode_loop+0x7d/0x80 do_syscall_64+0x18b/0x190 entry_SYSCALL_64_after_hwframe+0x3d/0xa2 The overflowed pgoff value causes hugetlbfs to try to set up a mapping with a negative range (end < start) that leaves invalid state which causes the BUG. The previous overflow fix to this code was incomplete and did not take the remap_file_pages system call into account. [mike.kravetz(a)oracle.com: v3] Link: http://lkml.kernel.org/r/20180309002726.7248-1-mike.kravetz@oracle.com [akpm(a)linux-foundation.org: include mmdebug.h] [akpm(a)linux-foundation.org: fix -ve left shift count on sh] Link: http://lkml.kernel.org/r/20180308210502.15952-1-mike.kravetz@oracle.com Fixes: 045c7a3f53d9 ("hugetlbfs: fix offset overflow in hugetlbfs mmap") Signed-off-by: Mike Kravetz <mike.kravetz(a)oracle.com> Reported-by: Nic Losby <blurbdust(a)gmail.com> Acked-by: Michal Hocko <mhocko(a)suse.com> Cc: "Kirill A . Shutemov" <kirill.shutemov(a)linux.intel.com> Cc: Yisheng Xie <xieyisheng1(a)huawei.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- fs/hugetlbfs/inode.c | 17 ++++++++++++++--- mm/hugetlb.c | 7 +++++++ 2 files changed, 21 insertions(+), 3 deletions(-) --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -118,6 +118,16 @@ static void huge_pagevec_release(struct pagevec_reinit(pvec); } +/* + * Mask used when checking the page offset value passed in via system + * calls. This value will be converted to a loff_t which is signed. + * Therefore, we want to check the upper PAGE_SHIFT + 1 bits of the + * value. The extra bit (- 1 in the shift value) is to take the sign + * bit into account. + */ +#define PGOFF_LOFFT_MAX \ + (((1UL << (PAGE_SHIFT + 1)) - 1) << (BITS_PER_LONG - (PAGE_SHIFT + 1))) + static int hugetlbfs_file_mmap(struct file *file, struct vm_area_struct *vma) { struct inode *inode = file_inode(file); @@ -137,12 +147,13 @@ static int hugetlbfs_file_mmap(struct fi vma->vm_ops = &hugetlb_vm_ops; /* - * Offset passed to mmap (before page shift) could have been - * negative when represented as a (l)off_t. + * page based offset in vm_pgoff could be sufficiently large to + * overflow a (l)off_t when converted to byte offset. */ - if (((loff_t)vma->vm_pgoff << PAGE_SHIFT) < 0) + if (vma->vm_pgoff & PGOFF_LOFFT_MAX) return -EINVAL; + /* must be huge page aligned */ if (vma->vm_pgoff & (~huge_page_mask(h) >> PAGE_SHIFT)) return -EINVAL; --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -18,6 +18,7 @@ #include <linux/bootmem.h> #include <linux/sysfs.h> #include <linux/slab.h> +#include <linux/mmdebug.h> #include <linux/sched/signal.h> #include <linux/rmap.h> #include <linux/string_helpers.h> @@ -4354,6 +4355,12 @@ int hugetlb_reserve_pages(struct inode * struct resv_map *resv_map; long gbl_reserve; + /* This should never happen */ + if (from > to) { + VM_WARN(1, "%s called with a negative range\n", __func__); + return -EINVAL; + } + /* * Only apply hugepage reservation if asked. At fault time, an * attempt will be made for VM_NORESERVE to allocate a page Patches currently in stable-queue which might be from mike.kravetz(a)oracle.com are queue-4.15/hugetlbfs-check-for-pgoff-value-overflow.patch

7 years, 9 months

1
0
0 0

Patch "h8300: remove extraneous __BIG_ENDIAN definition" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled h8300: remove extraneous __BIG_ENDIAN definition to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: h8300-remove-extraneous-__big_endian-definition.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 1705f7c534163594f8b05e060cb49fbea86ca70b Mon Sep 17 00:00:00 2001 From: Arnd Bergmann <arnd(a)arndb.de> Date: Thu, 22 Mar 2018 16:17:17 -0700 Subject: h8300: remove extraneous __BIG_ENDIAN definition From: Arnd Bergmann <arnd(a)arndb.de> commit 1705f7c534163594f8b05e060cb49fbea86ca70b upstream. A bugfix I did earlier caused a build regression on h8300, which defines the __BIG_ENDIAN macro in a slightly different way than the generic code: arch/h8300/include/asm/byteorder.h:5:0: warning: "__BIG_ENDIAN" redefined We don't need to define it here, as the same macro is already provided by the linux/byteorder/big_endian.h, and that version does not conflict. While this is a v4.16 regression, my earlier patch also got backported to the 4.14 and 4.15 stable kernels, so we need the fixup there as well. Link: http://lkml.kernel.org/r/20180313120752.2645129-1-arnd@arndb.de Fixes: 101110f6271c ("Kbuild: always define endianess in kconfig.h") Signed-off-by: Arnd Bergmann <arnd(a)arndb.de> Cc: Yoshinori Sato <ysato(a)users.sourceforge.jp> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/h8300/include/asm/byteorder.h | 1 - 1 file changed, 1 deletion(-) --- a/arch/h8300/include/asm/byteorder.h +++ b/arch/h8300/include/asm/byteorder.h @@ -2,7 +2,6 @@ #ifndef __H8300_BYTEORDER_H__ #define __H8300_BYTEORDER_H__ -#define __BIG_ENDIAN __ORDER_BIG_ENDIAN__ #include <linux/byteorder/big_endian.h> #endif Patches currently in stable-queue which might be from arnd(a)arndb.de are queue-4.15/h8300-remove-extraneous-__big_endian-definition.patch

7 years, 9 months

1
0
0 0

Patch "drm/vmwgfx: Fix black screen and device errors when running without fbdev" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/vmwgfx: Fix black screen and device errors when running without fbdev to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-vmwgfx-fix-black-screen-and-device-errors-when-running-without-fbdev.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 140bcaa23a1c37b694910424075a15e009120dbe Mon Sep 17 00:00:00 2001 From: Thomas Hellstrom <thellstrom(a)vmware.com> Date: Thu, 8 Mar 2018 10:07:37 +0100 Subject: drm/vmwgfx: Fix black screen and device errors when running without fbdev From: Thomas Hellstrom <thellstrom(a)vmware.com> commit 140bcaa23a1c37b694910424075a15e009120dbe upstream. When we are running without fbdev, transitioning from the login screen to X or gnome-shell/wayland will cause a vt switch and the driver will disable svga mode, losing all modesetting resources. However, the kms atomic state does not reflect that and may think that a crtc is still turned on, which will cause device errors when we try to bind an fb to the crtc, and the screen will remain black. Fix this by turning off all kms resources before disabling svga mode. Cc: <stable(a)vger.kernel.org> Signed-off-by: Thomas Hellstrom <thellstrom(a)vmware.com> Reviewed-by: Sinclair Yeh <syeh(a)vmware.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/vmwgfx/vmwgfx_drv.c | 13 +++++++++++++ drivers/gpu/drm/vmwgfx/vmwgfx_drv.h | 1 + drivers/gpu/drm/vmwgfx/vmwgfx_kms.c | 11 +++++++++++ drivers/gpu/drm/vmwgfx/vmwgfx_kms.h | 1 - 4 files changed, 25 insertions(+), 1 deletion(-) --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c @@ -1337,6 +1337,19 @@ static void __vmw_svga_disable(struct vm */ void vmw_svga_disable(struct vmw_private *dev_priv) { + /* + * Disabling SVGA will turn off device modesetting capabilities, so + * notify KMS about that so that it doesn't cache atomic state that + * isn't valid anymore, for example crtcs turned on. + * Strictly we'd want to do this under the SVGA lock (or an SVGA mutex), + * but vmw_kms_lost_device() takes the reservation sem and thus we'll + * end up with lock order reversal. Thus, a master may actually perform + * a new modeset just after we call vmw_kms_lost_device() and race with + * vmw_svga_disable(), but that should at worst cause atomic KMS state + * to be inconsistent with the device, causing modesetting problems. + * + */ + vmw_kms_lost_device(dev_priv->dev); ttm_write_lock(&dev_priv->reservation_sem, false); spin_lock(&dev_priv->svga_lock); if (dev_priv->bdev.man[TTM_PL_VRAM].use_type) { --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h @@ -938,6 +938,7 @@ int vmw_kms_present(struct vmw_private * int vmw_kms_update_layout_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv); void vmw_kms_legacy_hotspot_clear(struct vmw_private *dev_priv); +void vmw_kms_lost_device(struct drm_device *dev); int vmw_dumb_create(struct drm_file *file_priv, struct drm_device *dev, --- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c @@ -2865,3 +2865,14 @@ int vmw_kms_set_config(struct drm_mode_s return drm_atomic_helper_set_config(set, ctx); } + + +/** + * vmw_kms_lost_device - Notify kms that modesetting capabilities will be lost + * + * @dev: Pointer to the drm device + */ +void vmw_kms_lost_device(struct drm_device *dev) +{ + drm_atomic_helper_shutdown(dev); +} --- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.h @@ -439,5 +439,4 @@ int vmw_kms_stdu_dma(struct vmw_private int vmw_kms_set_config(struct drm_mode_set *set, struct drm_modeset_acquire_ctx *ctx); - #endif Patches currently in stable-queue which might be from thellstrom(a)vmware.com are queue-4.15/drm-vmwgfx-fix-a-destoy-while-held-mutex-problem.patch queue-4.15/drm-vmwgfx-fix-black-screen-and-device-errors-when-running-without-fbdev.patch

7 years, 9 months

1
0
0 0

Patch "drm/vmwgfx: Fix a destoy-while-held mutex problem." has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/vmwgfx: Fix a destoy-while-held mutex problem. to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-vmwgfx-fix-a-destoy-while-held-mutex-problem.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 73a88250b70954a8f27c2444e1c2411bba3c29d9 Mon Sep 17 00:00:00 2001 From: Thomas Hellstrom <thellstrom(a)vmware.com> Date: Wed, 21 Mar 2018 10:18:38 +0100 Subject: drm/vmwgfx: Fix a destoy-while-held mutex problem. From: Thomas Hellstrom <thellstrom(a)vmware.com> commit 73a88250b70954a8f27c2444e1c2411bba3c29d9 upstream. When validating legacy surfaces, the backup bo might be destroyed at surface validate time. However, the kms resource validation code may have the bo reserved, so we will destroy a locked mutex. While there shouldn't be any other users of that mutex when it is destroyed, it causes a lock leak and thus throws a lockdep error. Fix this by having the kms resource validation code hold a reference to the bo while we have it reserved. We do this by introducing a validation context which might come in handy when the kms code is extended to validate multiple resources or buffers. Cc: <stable(a)vger.kernel.org> Signed-off-by: Thomas Hellstrom <thellstrom(a)vmware.com> Reviewed-by: Brian Paul <brianp(a)vmware.com> Reviewed-by: Sinclair Yeh <syeh(a)vmware.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/vmwgfx/vmwgfx_kms.c | 28 +++++++++++++++++++--------- drivers/gpu/drm/vmwgfx/vmwgfx_kms.h | 12 +++++++++--- drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c | 5 +++-- drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c | 5 +++-- 4 files changed, 34 insertions(+), 16 deletions(-) --- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c @@ -31,7 +31,6 @@ #include <drm/drm_atomic_helper.h> #include <drm/drm_rect.h> - /* Might need a hrtimer here? */ #define VMWGFX_PRESENT_RATE ((HZ / 60 > 0) ? HZ / 60 : 1) @@ -2531,9 +2530,12 @@ void vmw_kms_helper_buffer_finish(struct * Helper to be used if an error forces the caller to undo the actions of * vmw_kms_helper_resource_prepare. */ -void vmw_kms_helper_resource_revert(struct vmw_resource *res) +void vmw_kms_helper_resource_revert(struct vmw_validation_ctx *ctx) { - vmw_kms_helper_buffer_revert(res->backup); + struct vmw_resource *res = ctx->res; + + vmw_kms_helper_buffer_revert(ctx->buf); + vmw_dmabuf_unreference(&ctx->buf); vmw_resource_unreserve(res, false, NULL, 0); mutex_unlock(&res->dev_priv->cmdbuf_mutex); } @@ -2550,10 +2552,14 @@ void vmw_kms_helper_resource_revert(stru * interrupted by a signal. */ int vmw_kms_helper_resource_prepare(struct vmw_resource *res, - bool interruptible) + bool interruptible, + struct vmw_validation_ctx *ctx) { int ret = 0; + ctx->buf = NULL; + ctx->res = res; + if (interruptible) ret = mutex_lock_interruptible(&res->dev_priv->cmdbuf_mutex); else @@ -2572,6 +2578,8 @@ int vmw_kms_helper_resource_prepare(stru res->dev_priv->has_mob); if (ret) goto out_unreserve; + + ctx->buf = vmw_dmabuf_reference(res->backup); } ret = vmw_resource_validate(res); if (ret) @@ -2579,7 +2587,7 @@ int vmw_kms_helper_resource_prepare(stru return 0; out_revert: - vmw_kms_helper_buffer_revert(res->backup); + vmw_kms_helper_buffer_revert(ctx->buf); out_unreserve: vmw_resource_unreserve(res, false, NULL, 0); out_unlock: @@ -2595,11 +2603,13 @@ out_unlock: * @out_fence: Optional pointer to a fence pointer. If non-NULL, a * ref-counted fence pointer is returned here. */ -void vmw_kms_helper_resource_finish(struct vmw_resource *res, - struct vmw_fence_obj **out_fence) +void vmw_kms_helper_resource_finish(struct vmw_validation_ctx *ctx, + struct vmw_fence_obj **out_fence) { - if (res->backup || out_fence) - vmw_kms_helper_buffer_finish(res->dev_priv, NULL, res->backup, + struct vmw_resource *res = ctx->res; + + if (ctx->buf || out_fence) + vmw_kms_helper_buffer_finish(res->dev_priv, NULL, ctx->buf, out_fence, NULL); vmw_resource_unreserve(res, false, NULL, 0); --- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.h @@ -240,6 +240,11 @@ struct vmw_display_unit { int set_gui_y; }; +struct vmw_validation_ctx { + struct vmw_resource *res; + struct vmw_dma_buffer *buf; +}; + #define vmw_crtc_to_du(x) \ container_of(x, struct vmw_display_unit, crtc) #define vmw_connector_to_du(x) \ @@ -296,9 +301,10 @@ void vmw_kms_helper_buffer_finish(struct struct drm_vmw_fence_rep __user * user_fence_rep); int vmw_kms_helper_resource_prepare(struct vmw_resource *res, - bool interruptible); -void vmw_kms_helper_resource_revert(struct vmw_resource *res); -void vmw_kms_helper_resource_finish(struct vmw_resource *res, + bool interruptible, + struct vmw_validation_ctx *ctx); +void vmw_kms_helper_resource_revert(struct vmw_validation_ctx *ctx); +void vmw_kms_helper_resource_finish(struct vmw_validation_ctx *ctx, struct vmw_fence_obj **out_fence); int vmw_kms_readback(struct vmw_private *dev_priv, struct drm_file *file_priv, --- a/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c @@ -909,12 +909,13 @@ int vmw_kms_sou_do_surface_dirty(struct struct vmw_framebuffer_surface *vfbs = container_of(framebuffer, typeof(*vfbs), base); struct vmw_kms_sou_surface_dirty sdirty; + struct vmw_validation_ctx ctx; int ret; if (!srf) srf = &vfbs->surface->res; - ret = vmw_kms_helper_resource_prepare(srf, true); + ret = vmw_kms_helper_resource_prepare(srf, true, &ctx); if (ret) return ret; @@ -933,7 +934,7 @@ int vmw_kms_sou_do_surface_dirty(struct ret = vmw_kms_helper_dirty(dev_priv, framebuffer, clips, vclips, dest_x, dest_y, num_clips, inc, &sdirty.base); - vmw_kms_helper_resource_finish(srf, out_fence); + vmw_kms_helper_resource_finish(&ctx, out_fence); return ret; } --- a/drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c @@ -980,12 +980,13 @@ int vmw_kms_stdu_surface_dirty(struct vm struct vmw_framebuffer_surface *vfbs = container_of(framebuffer, typeof(*vfbs), base); struct vmw_stdu_dirty sdirty; + struct vmw_validation_ctx ctx; int ret; if (!srf) srf = &vfbs->surface->res; - ret = vmw_kms_helper_resource_prepare(srf, true); + ret = vmw_kms_helper_resource_prepare(srf, true, &ctx); if (ret) return ret; @@ -1008,7 +1009,7 @@ int vmw_kms_stdu_surface_dirty(struct vm dest_x, dest_y, num_clips, inc, &sdirty.base); out_finish: - vmw_kms_helper_resource_finish(srf, out_fence); + vmw_kms_helper_resource_finish(&ctx, out_fence); return ret; } Patches currently in stable-queue which might be from thellstrom(a)vmware.com are queue-4.15/drm-vmwgfx-fix-a-destoy-while-held-mutex-problem.patch queue-4.15/drm-vmwgfx-fix-black-screen-and-device-errors-when-running-without-fbdev.patch

7 years, 9 months

1
0
0 0

Patch "drm: udl: Properly check framebuffer mmap offsets" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm: udl: Properly check framebuffer mmap offsets to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-udl-properly-check-framebuffer-mmap-offsets.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 3b82a4db8eaccce735dffd50b4d4e1578099b8e8 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Date: Wed, 21 Mar 2018 16:45:53 +0100 Subject: drm: udl: Properly check framebuffer mmap offsets From: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> commit 3b82a4db8eaccce735dffd50b4d4e1578099b8e8 upstream. The memmap options sent to the udl framebuffer driver were not being checked for all sets of possible crazy values. Fix this up by properly bounding the allowed values. Reported-by: Eyal Itkin <eyalit(a)checkpoint.com> Cc: stable <stable(a)vger.kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Daniel Vetter <daniel.vetter(a)ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20180321154553.GA18454@kroah.… Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/udl/udl_fb.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) --- a/drivers/gpu/drm/udl/udl_fb.c +++ b/drivers/gpu/drm/udl/udl_fb.c @@ -159,10 +159,15 @@ static int udl_fb_mmap(struct fb_info *i { unsigned long start = vma->vm_start; unsigned long size = vma->vm_end - vma->vm_start; - unsigned long offset = vma->vm_pgoff << PAGE_SHIFT; + unsigned long offset; unsigned long page, pos; - if (offset + size > info->fix.smem_len) + if (vma->vm_pgoff > (~0UL >> PAGE_SHIFT)) + return -EINVAL; + + offset = vma->vm_pgoff << PAGE_SHIFT; + + if (offset > info->fix.smem_len || size > info->fix.smem_len - offset) return -EINVAL; pos = (unsigned long)info->fix.smem_start + offset; Patches currently in stable-queue which might be from gregkh(a)linuxfoundation.org are queue-4.15/mips-lantiq-ase-enable-mfd_syscon.patch queue-4.15/mm-khugepaged.c-convert-vm_bug_on-to-collapse-fail.patch queue-4.15/drm-vmwgfx-fix-a-destoy-while-held-mutex-problem.patch queue-4.15/mmc-core-disable-hpi-for-certain-micron-numonyx-emmc-cards.patch queue-4.15/alsa-aloop-fix-access-to-not-yet-ready-substream-via-cable.patch queue-4.15/alsa-hda-realtek-fix-speaker-no-sound-after-system-resume.patch queue-4.15/libata-disable-lpm-for-crucial-bx100-ssd-500gb-drive.patch queue-4.15/can-ifi-check-core-revision-upon-probe.patch queue-4.15/bluetooth-btusb-fix-quirk-for-atheros-1525-qca6174.patch queue-4.15/nfsd-remove-blocked-locks-on-client-teardown.patch queue-4.15/mtdchar-fix-usage-of-mtd_ooblayout_ecc.patch queue-4.15/libata-apply-nolpm-quirk-to-crucial-m500-480-and-960gb-ssds.patch queue-4.15/drm-radeon-don-t-turn-off-dp-sink-when-disconnected.patch queue-4.15/clk-bcm2835-protect-sections-updating-shared-registers.patch queue-4.15/alsa-hda-realtek-fix-dell-headset-mic-can-t-record.patch queue-4.15/mmc-dw_mmc-fix-the-dto-cto-timeout-overflow-calculation-for-32-bit-systems.patch queue-4.15/sched-cgroup-don-t-reject-lower-cpu.max-on-ancestors.patch queue-4.15/module-propagate-error-in-modules_open.patch queue-4.15/rtlwifi-rtl8723be-fix-loss-of-signal.patch queue-4.15/libata-fix-length-validation-of-atapi-relayed-scsi-commands.patch queue-4.15/can-peak-pcie_fd-fix-echo_skb-is-occupied-bug.patch queue-4.15/tracing-probeevent-fix-to-support-minus-offset-from-symbol.patch queue-4.15/can-cc770-fix-use-after-free-in-cc770_tx_interrupt.patch queue-4.15/rdma-mlx5-fix-crash-while-accessing-garbage-pointer-and-freed-memory.patch queue-4.15/libata-apply-nolpm-quirk-to-crucial-mx100-512gb-ssds.patch queue-4.15/alsa-hda-realtek-always-immediately-update-mute-led-with-pin-vref.patch queue-4.15/iio-st_pressure-st_accel-pass-correct-platform-data-to-init.patch queue-4.15/h8300-remove-extraneous-__big_endian-definition.patch queue-4.15/iio-chemical-ccs811-corrected-firmware-boot-application-mode-transition.patch queue-4.15/pinctrl-samsung-validate-alias-coming-from-dt.patch queue-4.15/pci-add-function-1-dma-alias-quirk-for-highpoint-rocketraid-644l.patch queue-4.15/revert-mm-page_alloc-skip-over-regions-of-invalid-pfns-where-possible.patch queue-4.15/libata-modify-quirks-for-mx100-to-limit-ncq_trim-quirk-to-mu01-version.patch queue-4.15/mtd-nand-fsl_ifc-fix-nand-waitfunc-return-value.patch queue-4.15/drivers-hv-vmbus-fix-ring-buffer-signaling.patch queue-4.15/drm-reject-getfb-for-multi-plane-framebuffers.patch queue-4.15/can-ifi-repair-the-error-handling.patch queue-4.15/can-cc770-fix-queue-stall-dropped-rtr-reply.patch queue-4.15/libata-enable-queued-trim-for-samsung-ssd-860.patch queue-4.15/staging-ncpfs-memory-corruption-in-ncp_read_kernel.patch queue-4.15/ahci-add-pci-id-for-the-highpoint-rocketraid-644l-card.patch queue-4.15/mmc-dw_mmc-exynos-fix-the-suspend-resume-issue-for-exynos5433.patch queue-4.15/drm-udl-properly-check-framebuffer-mmap-offsets.patch queue-4.15/mips-lantiq-enable-ahb-bus-for-usb.patch queue-4.15/can-cc770-fix-stalls-on-rt-linux-remove-redundant-irq-ack.patch queue-4.15/iio-adc-meson-saradc-unlock-on-error-in-meson_sar_adc_lock.patch queue-4.15/mips-ralink-fix-booting-on-mt7621.patch queue-4.15/clk-sunxi-ng-a31-fix-clk_out_-clock-ops.patch queue-4.15/media-tegra-cec-reset-rx_buf_cnt-when-start-bit-detected.patch queue-4.15/brcmfmac-fix-p2p_device-ethernet-address-generation.patch queue-4.15/mm-thp-do-not-wait-for-lock_page-in-deferred_split_scan.patch queue-4.15/mtd-nand-fsl_ifc-fix-eccstat-array-overflow-for-ifc-ver-2.0.0.patch queue-4.15/hugetlbfs-check-for-pgoff-value-overflow.patch queue-4.15/lockdep-fix-fs_reclaim-warning.patch queue-4.15/mips-lantiq-fix-danube-usb-clock.patch queue-4.15/cgroup-fix-rule-checking-for-threaded-mode-switching.patch queue-4.15/mmc-core-fix-tracepoint-print-of-blk_addr-and-blksz.patch queue-4.15/alsa-usb-audio-fix-parsing-descriptor-of-uac2-processing-unit.patch queue-4.15/mm-vmscan-wake-up-flushers-for-legacy-cgroups-too.patch queue-4.15/acpi-numa-fix-pxm-to-online-numa-node-associations.patch queue-4.15/mmc-block-fix-updating-ext_csd-caches-on-ioctl-call.patch queue-4.15/drm-amd-display-add-one-to-edid-s-audio-channel-count-when-passing-to-dc.patch queue-4.15/can-peak-pcie_fd-remove-useless-code-when-interface-starts.patch queue-4.15/alsa-aloop-sync-stale-timer-before-release.patch queue-4.15/drm-amd-display-we-shouldn-t-set-format_default-on-plane-as-atomic-driver.patch queue-4.15/mips-ralink-remove-ralink_halt.patch queue-4.15/libnvdimm-btt-blk-do-integrity-setup-before-add_disk.patch queue-4.15/acpi-watchdog-fix-off-by-one-error-at-resource-assignment.patch queue-4.15/mtd-nand-fsl_ifc-read-eccstat0-and-eccstat1-registers-for-ifc-2.0.patch queue-4.15/mmc-dw_mmc-fix-falling-from-idmac-to-pio-mode-when-dw_mci_reset-occurs.patch queue-4.15/libata-make-crucial-bx100-500gb-lpm-quirk-apply-to-all-firmware-versions.patch queue-4.15/clk-bcm2835-fix-ana-maskx-definitions.patch queue-4.15/alsa-hda-force-polling-mode-on-cfl-for-fixing-codec-communication.patch queue-4.15/libata-remove-warn-for-dma-or-pio-command-without-data.patch queue-4.15/drm-vmwgfx-fix-black-screen-and-device-errors-when-running-without-fbdev.patch queue-4.15/mm-vmalloc-add-interfaces-to-free-unmapped-page-table.patch queue-4.15/bluetooth-btusb-add-dell-optiplex-3060-to-btusb_needs_reset_resume_table.patch queue-4.15/mm-shmem-do-not-wait-for-lock_page-in-shmem_unused_huge_shrink.patch queue-4.15/bluetooth-btusb-remove-yoga-920-from-the-btusb_needs_reset_resume_table.patch queue-4.15/x86-mm-implement-free-pmd-pte-page-interfaces.patch queue-4.15/libata-don-t-try-to-pass-through-ncq-commands-to-non-ncq-devices.patch

7 years, 9 months

1
0
0 0

Patch "drm: Reject getfb for multi-plane framebuffers" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm: Reject getfb for multi-plane framebuffers to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-reject-getfb-for-multi-plane-framebuffers.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From b24791fe00f8b089d5b10cb7bcc4e1ae88b4831b Mon Sep 17 00:00:00 2001 From: Daniel Stone <daniels(a)collabora.com> Date: Tue, 20 Mar 2018 22:58:39 +0000 Subject: drm: Reject getfb for multi-plane framebuffers From: Daniel Stone <daniels(a)collabora.com> commit b24791fe00f8b089d5b10cb7bcc4e1ae88b4831b upstream. getfb can only return a single plane, so reject attempts to use it with multi-plane framebuffers. Signed-off-by: Daniel Stone <daniels(a)collabora.com> Reported-by: Daniel van Vugt <daniel.van.vugt(a)canonical.com> Reviewed-by: Rob Clark <robdclark(a)gmail.com> Reviewed-by: Daniel Vetter <daniel.vetter(a)ffwll.ch> Fixes: 308e5bcbdb10 ("drm: add an fb creation ioctl that takes a pixel format v5") Cc: stable(a)vger.kernel.org # v3.3+ Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105518 Link: https://patchwork.freedesktop.org/patch/msgid/20180320225839.30905-1-daniel… Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/drm_framebuffer.c | 7 +++++++ 1 file changed, 7 insertions(+) --- a/drivers/gpu/drm/drm_framebuffer.c +++ b/drivers/gpu/drm/drm_framebuffer.c @@ -458,6 +458,12 @@ int drm_mode_getfb(struct drm_device *de if (!fb) return -ENOENT; + /* Multi-planar framebuffers need getfb2. */ + if (fb->format->num_planes > 1) { + ret = -EINVAL; + goto out; + } + r->height = fb->height; r->width = fb->width; r->depth = fb->format->depth; @@ -481,6 +487,7 @@ int drm_mode_getfb(struct drm_device *de ret = -ENODEV; } +out: drm_framebuffer_put(fb); return ret; Patches currently in stable-queue which might be from daniels(a)collabora.com are queue-4.15/drm-reject-getfb-for-multi-plane-framebuffers.patch

7 years, 9 months

1
0
0 0

Patch "drm/radeon: Don't turn off DP sink when disconnected" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/radeon: Don't turn off DP sink when disconnected to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-radeon-don-t-turn-off-dp-sink-when-disconnected.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 2681bc79eeb640562c932007bfebbbdc55bf6a7d Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Michel=20D=C3=A4nzer?= <michel.daenzer(a)amd.com> Date: Wed, 14 Mar 2018 18:14:04 +0100 Subject: drm/radeon: Don't turn off DP sink when disconnected MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit From: Michel Dänzer <michel.daenzer(a)amd.com> commit 2681bc79eeb640562c932007bfebbbdc55bf6a7d upstream. Turning off the sink in this case causes various issues, because userspace expects it to stay on until it turns it off explicitly. Instead, turn the sink off and back on when a display is connected again. This dance seems necessary for link training to work correctly. Bugzilla: https://bugs.freedesktop.org/105308 Cc: stable(a)vger.kernel.org Reviewed-by: Alex Deucher <alexander.deucher(a)amd.com> Signed-off-by: Michel Dänzer <michel.daenzer(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/radeon/radeon_connectors.c | 29 +++++++++++------------------ 1 file changed, 11 insertions(+), 18 deletions(-) --- a/drivers/gpu/drm/radeon/radeon_connectors.c +++ b/drivers/gpu/drm/radeon/radeon_connectors.c @@ -90,25 +90,18 @@ void radeon_connector_hotplug(struct drm /* don't do anything if sink is not display port, i.e., * passive dp->(dvi|hdmi) adaptor */ - if (dig_connector->dp_sink_type == CONNECTOR_OBJECT_ID_DISPLAYPORT) { - int saved_dpms = connector->dpms; - /* Only turn off the display if it's physically disconnected */ - if (!radeon_hpd_sense(rdev, radeon_connector->hpd.hpd)) { - drm_helper_connector_dpms(connector, DRM_MODE_DPMS_OFF); - } else if (radeon_dp_needs_link_train(radeon_connector)) { - /* Don't try to start link training before we - * have the dpcd */ - if (!radeon_dp_getdpcd(radeon_connector)) - return; + if (dig_connector->dp_sink_type == CONNECTOR_OBJECT_ID_DISPLAYPORT && + radeon_hpd_sense(rdev, radeon_connector->hpd.hpd) && + radeon_dp_needs_link_train(radeon_connector)) { + /* Don't start link training before we have the DPCD */ + if (!radeon_dp_getdpcd(radeon_connector)) + return; - /* set it to OFF so that drm_helper_connector_dpms() - * won't return immediately since the current state - * is ON at this point. - */ - connector->dpms = DRM_MODE_DPMS_OFF; - drm_helper_connector_dpms(connector, DRM_MODE_DPMS_ON); - } - connector->dpms = saved_dpms; + /* Turn the connector off and back on immediately, which + * will trigger link training + */ + drm_helper_connector_dpms(connector, DRM_MODE_DPMS_OFF); + drm_helper_connector_dpms(connector, DRM_MODE_DPMS_ON); } } } Patches currently in stable-queue which might be from michel.daenzer(a)amd.com are queue-4.15/drm-radeon-don-t-turn-off-dp-sink-when-disconnected.patch

7 years, 9 months

1
0
0 0

Patch "drm/amd/display: We shouldn't set format_default on plane as atomic driver" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/amd/display: We shouldn't set format_default on plane as atomic driver to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-amd-display-we-shouldn-t-set-format_default-on-plane-as-atomic-driver.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 509648fcf0ce8650184649b43ad039f78dde155f Mon Sep 17 00:00:00 2001 From: Harry Wentland <harry.wentland(a)amd.com> Date: Tue, 6 Mar 2018 11:14:12 -0500 Subject: drm/amd/display: We shouldn't set format_default on plane as atomic driver From: Harry Wentland <harry.wentland(a)amd.com> commit 509648fcf0ce8650184649b43ad039f78dde155f upstream. This is still a leftover from early atomic brinup days. Signed-off-by: Harry Wentland <harry.wentland(a)amd.com> Reviewed-by: Tony Cheng <Tony.Cheng(a)amd.com> Acked-by: Harry Wentland <harry.wentland(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> Cc: stable(a)vger.kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 2 -- 1 file changed, 2 deletions(-) --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -3090,8 +3090,6 @@ static int amdgpu_dm_plane_init(struct a switch (aplane->base.type) { case DRM_PLANE_TYPE_PRIMARY: - aplane->base.format_default = true; - res = drm_universal_plane_init( dm->adev->ddev, &aplane->base, Patches currently in stable-queue which might be from harry.wentland(a)amd.com are queue-4.15/drm-amd-display-add-one-to-edid-s-audio-channel-count-when-passing-to-dc.patch queue-4.15/drm-amd-display-we-shouldn-t-set-format_default-on-plane-as-atomic-driver.patch

7 years, 9 months

1
0
0 0

Patch "drm/amd/display: Add one to EDID's audio channel count when passing to DC" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/amd/display: Add one to EDID's audio channel count when passing to DC to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-amd-display-add-one-to-edid-s-audio-channel-count-when-passing-to-dc.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 731a373698c9675d5aed8a30d8c9861bea9c41a2 Mon Sep 17 00:00:00 2001 From: Harry Wentland <harry.wentland(a)amd.com> Date: Wed, 7 Mar 2018 13:45:33 -0500 Subject: drm/amd/display: Add one to EDID's audio channel count when passing to DC From: Harry Wentland <harry.wentland(a)amd.com> commit 731a373698c9675d5aed8a30d8c9861bea9c41a2 upstream. DC takes channel count to mean the actual count. cea_sad's channels represent it as number of channels - 1. Signed-off-by: Harry Wentland <harry.wentland(a)amd.com> Reviewed-by: Tony Cheng <Tony.Cheng(a)amd.com> Acked-by: Harry Wentland <harry.wentland(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> Cc: stable(a)vger.kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c @@ -109,7 +109,7 @@ enum dc_edid_status dm_helpers_parse_edi struct cea_sad *sad = &sads[i]; edid_caps->audio_modes[i].format_code = sad->format; - edid_caps->audio_modes[i].channel_count = sad->channels; + edid_caps->audio_modes[i].channel_count = sad->channels + 1; edid_caps->audio_modes[i].sample_rate = sad->freq; edid_caps->audio_modes[i].sample_size = sad->byte2; } Patches currently in stable-queue which might be from harry.wentland(a)amd.com are queue-4.15/drm-amd-display-add-one-to-edid-s-audio-channel-count-when-passing-to-dc.patch queue-4.15/drm-amd-display-we-shouldn-t-set-format_default-on-plane-as-atomic-driver.patch

7 years, 9 months

1
0
0 0

Patch "can: peak/pcie_fd: remove useless code when interface starts" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled can: peak/pcie_fd: remove useless code when interface starts to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: can-peak-pcie_fd-remove-useless-code-when-interface-starts.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From ffd137f7043cb30067e1bff6fe62a073ae190b23 Mon Sep 17 00:00:00 2001 From: Stephane Grosjean <s.grosjean(a)peak-system.com> Date: Thu, 8 Mar 2018 09:30:29 +0100 Subject: can: peak/pcie_fd: remove useless code when interface starts From: Stephane Grosjean <s.grosjean(a)peak-system.com> commit ffd137f7043cb30067e1bff6fe62a073ae190b23 upstream. When an interface starts, the echo_skb array is empty and the network queue should be started only. This patch replaces useless code and locks when the internal RX_BARRIER message is received from the IP core, telling the driver that tx may start. Signed-off-by: Stephane Grosjean <s.grosjean(a)peak-system.com> Cc: linux-stable <stable(a)vger.kernel.org> Signed-off-by: Marc Kleine-Budde <mkl(a)pengutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/can/peak_canfd/peak_canfd.c | 13 ++----------- 1 file changed, 2 insertions(+), 11 deletions(-) --- a/drivers/net/can/peak_canfd/peak_canfd.c +++ b/drivers/net/can/peak_canfd/peak_canfd.c @@ -333,7 +333,6 @@ static int pucan_handle_status(struct pe /* this STATUS is the CNF of the RX_BARRIER: Tx path can be setup */ if (pucan_status_is_rx_barrier(msg)) { - unsigned long flags; if (priv->enable_tx_path) { int err = priv->enable_tx_path(priv); @@ -342,16 +341,8 @@ static int pucan_handle_status(struct pe return err; } - /* restart network queue only if echo skb array is free */ - spin_lock_irqsave(&priv->echo_lock, flags); - - if (!priv->can.echo_skb[priv->echo_idx]) { - spin_unlock_irqrestore(&priv->echo_lock, flags); - - netif_wake_queue(ndev); - } else { - spin_unlock_irqrestore(&priv->echo_lock, flags); - } + /* start network queue (echo_skb array is empty) */ + netif_start_queue(ndev); return 0; } Patches currently in stable-queue which might be from s.grosjean(a)peak-system.com are queue-4.15/can-peak-pcie_fd-fix-echo_skb-is-occupied-bug.patch queue-4.15/can-peak-pcie_fd-remove-useless-code-when-interface-starts.patch

7 years, 9 months

1
0
0 0

Patch "can: peak/pcie_fd: fix echo_skb is occupied! bug" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled can: peak/pcie_fd: fix echo_skb is occupied! bug to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: can-peak-pcie_fd-fix-echo_skb-is-occupied-bug.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From e6048a00cfd0863d32f53b226e0b9a3633fc3332 Mon Sep 17 00:00:00 2001 From: Stephane Grosjean <s.grosjean(a)peak-system.com> Date: Thu, 8 Mar 2018 09:30:28 +0100 Subject: can: peak/pcie_fd: fix echo_skb is occupied! bug From: Stephane Grosjean <s.grosjean(a)peak-system.com> commit e6048a00cfd0863d32f53b226e0b9a3633fc3332 upstream. This patch makes atomic the handling of the linux-can echo_skb array and the network tx queue. This prevents from the "BUG! echo_skb is occupied!" message to be printed by the linux-can core, in SMP environments. Reported-by: Diana Burgess <diana(a)peloton-tech.com> Signed-off-by: Stephane Grosjean <s.grosjean(a)peak-system.com> Cc: linux-stable <stable(a)vger.kernel.org> Signed-off-by: Marc Kleine-Budde <mkl(a)pengutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/can/peak_canfd/peak_canfd.c | 12 ++++++------ drivers/net/can/peak_canfd/peak_pciefd_main.c | 8 ++++++-- 2 files changed, 12 insertions(+), 8 deletions(-) --- a/drivers/net/can/peak_canfd/peak_canfd.c +++ b/drivers/net/can/peak_canfd/peak_canfd.c @@ -262,7 +262,6 @@ static int pucan_handle_can_rx(struct pe spin_lock_irqsave(&priv->echo_lock, flags); can_get_echo_skb(priv->ndev, msg->client); - spin_unlock_irqrestore(&priv->echo_lock, flags); /* count bytes of the echo instead of skb */ stats->tx_bytes += cf_len; @@ -271,6 +270,7 @@ static int pucan_handle_can_rx(struct pe /* restart tx queue (a slot is free) */ netif_wake_queue(priv->ndev); + spin_unlock_irqrestore(&priv->echo_lock, flags); return 0; } @@ -726,11 +726,6 @@ static netdev_tx_t peak_canfd_start_xmit */ should_stop_tx_queue = !!(priv->can.echo_skb[priv->echo_idx]); - spin_unlock_irqrestore(&priv->echo_lock, flags); - - /* write the skb on the interface */ - priv->write_tx_msg(priv, msg); - /* stop network tx queue if not enough room to save one more msg too */ if (priv->can.ctrlmode & CAN_CTRLMODE_FD) should_stop_tx_queue |= (room_left < @@ -742,6 +737,11 @@ static netdev_tx_t peak_canfd_start_xmit if (should_stop_tx_queue) netif_stop_queue(ndev); + spin_unlock_irqrestore(&priv->echo_lock, flags); + + /* write the skb on the interface */ + priv->write_tx_msg(priv, msg); + return NETDEV_TX_OK; } --- a/drivers/net/can/peak_canfd/peak_pciefd_main.c +++ b/drivers/net/can/peak_canfd/peak_pciefd_main.c @@ -349,8 +349,12 @@ static irqreturn_t pciefd_irq_handler(in priv->tx_pages_free++; spin_unlock_irqrestore(&priv->tx_lock, flags); - /* wake producer up */ - netif_wake_queue(priv->ucan.ndev); + /* wake producer up (only if enough room in echo_skb array) */ + spin_lock_irqsave(&priv->ucan.echo_lock, flags); + if (!priv->ucan.can.echo_skb[priv->ucan.echo_idx]) + netif_wake_queue(priv->ucan.ndev); + + spin_unlock_irqrestore(&priv->ucan.echo_lock, flags); } /* re-enable Rx DMA transfer for this CAN */ Patches currently in stable-queue which might be from s.grosjean(a)peak-system.com are queue-4.15/can-peak-pcie_fd-fix-echo_skb-is-occupied-bug.patch queue-4.15/can-peak-pcie_fd-remove-useless-code-when-interface-starts.patch

7 years, 9 months

1
0
0 0

Patch "can: ifi: Repair the error handling" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled can: ifi: Repair the error handling to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: can-ifi-repair-the-error-handling.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 880dd464b4304583c557c4e5f5ecebfd55d232b1 Mon Sep 17 00:00:00 2001 From: Marek Vasut <marex(a)denx.de> Date: Thu, 1 Mar 2018 19:34:00 +0100 Subject: can: ifi: Repair the error handling From: Marek Vasut <marex(a)denx.de> commit 880dd464b4304583c557c4e5f5ecebfd55d232b1 upstream. The new version of the IFI CANFD core has significantly less complex error state indication logic. In particular, the warning/error state bits are no longer all over the place, but are all present in the STATUS register. Moreover, there is a new IRQ register bit indicating transition between error states (active/warning/passive/busoff). This patch makes use of this bit to weed out the obscure selective INTERRUPT register clearing, which was used to carry over the error state indication into the poll function. While at it, this patch fixes the handling of the ACTIVE state, since the hardware provides indication of the core being in ACTIVE state and that in turn fixes the state transition indication toward userspace. Finally, register reads in the poll function are moved to the matching subfunctions since those are also no longer needed in the poll function. Signed-off-by: Marek Vasut <marex(a)denx.de> Cc: Heiko Schocher <hs(a)denx.de> Cc: Markus Marb <markus(a)marb.org> Cc: Marc Kleine-Budde <mkl(a)pengutronix.de> Cc: linux-stable <stable(a)vger.kernel.org> Signed-off-by: Marc Kleine-Budde <mkl(a)pengutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/can/ifi_canfd/ifi_canfd.c | 64 +++++++++++++++++++--------------- 1 file changed, 37 insertions(+), 27 deletions(-) --- a/drivers/net/can/ifi_canfd/ifi_canfd.c +++ b/drivers/net/can/ifi_canfd/ifi_canfd.c @@ -30,6 +30,7 @@ #define IFI_CANFD_STCMD_ERROR_ACTIVE BIT(2) #define IFI_CANFD_STCMD_ERROR_PASSIVE BIT(3) #define IFI_CANFD_STCMD_BUSOFF BIT(4) +#define IFI_CANFD_STCMD_ERROR_WARNING BIT(5) #define IFI_CANFD_STCMD_BUSMONITOR BIT(16) #define IFI_CANFD_STCMD_LOOPBACK BIT(18) #define IFI_CANFD_STCMD_DISABLE_CANFD BIT(24) @@ -52,7 +53,10 @@ #define IFI_CANFD_TXSTCMD_OVERFLOW BIT(13) #define IFI_CANFD_INTERRUPT 0xc +#define IFI_CANFD_INTERRUPT_ERROR_BUSOFF BIT(0) #define IFI_CANFD_INTERRUPT_ERROR_WARNING BIT(1) +#define IFI_CANFD_INTERRUPT_ERROR_STATE_CHG BIT(2) +#define IFI_CANFD_INTERRUPT_ERROR_REC_TEC_INC BIT(3) #define IFI_CANFD_INTERRUPT_ERROR_COUNTER BIT(10) #define IFI_CANFD_INTERRUPT_TXFIFO_EMPTY BIT(16) #define IFI_CANFD_INTERRUPT_TXFIFO_REMOVE BIT(22) @@ -61,6 +65,10 @@ #define IFI_CANFD_INTERRUPT_SET_IRQ ((u32)BIT(31)) #define IFI_CANFD_IRQMASK 0x10 +#define IFI_CANFD_IRQMASK_ERROR_BUSOFF BIT(0) +#define IFI_CANFD_IRQMASK_ERROR_WARNING BIT(1) +#define IFI_CANFD_IRQMASK_ERROR_STATE_CHG BIT(2) +#define IFI_CANFD_IRQMASK_ERROR_REC_TEC_INC BIT(3) #define IFI_CANFD_IRQMASK_SET_ERR BIT(7) #define IFI_CANFD_IRQMASK_SET_TS BIT(15) #define IFI_CANFD_IRQMASK_TXFIFO_EMPTY BIT(16) @@ -220,7 +228,10 @@ static void ifi_canfd_irq_enable(struct if (enable) { enirq = IFI_CANFD_IRQMASK_TXFIFO_EMPTY | - IFI_CANFD_IRQMASK_RXFIFO_NEMPTY; + IFI_CANFD_IRQMASK_RXFIFO_NEMPTY | + IFI_CANFD_IRQMASK_ERROR_STATE_CHG | + IFI_CANFD_IRQMASK_ERROR_WARNING | + IFI_CANFD_IRQMASK_ERROR_BUSOFF; if (priv->can.ctrlmode & CAN_CTRLMODE_BERR_REPORTING) enirq |= IFI_CANFD_INTERRUPT_ERROR_COUNTER; } @@ -361,12 +372,13 @@ static int ifi_canfd_handle_lost_msg(str return 1; } -static int ifi_canfd_handle_lec_err(struct net_device *ndev, const u32 errctr) +static int ifi_canfd_handle_lec_err(struct net_device *ndev) { struct ifi_canfd_priv *priv = netdev_priv(ndev); struct net_device_stats *stats = &ndev->stats; struct can_frame *cf; struct sk_buff *skb; + u32 errctr = readl(priv->base + IFI_CANFD_ERROR_CTR); const u32 errmask = IFI_CANFD_ERROR_CTR_OVERLOAD_FIRST | IFI_CANFD_ERROR_CTR_ACK_ERROR_FIRST | IFI_CANFD_ERROR_CTR_BIT0_ERROR_FIRST | @@ -449,6 +461,11 @@ static int ifi_canfd_handle_state_change switch (new_state) { case CAN_STATE_ERROR_ACTIVE: + /* error active state */ + priv->can.can_stats.error_warning++; + priv->can.state = CAN_STATE_ERROR_ACTIVE; + break; + case CAN_STATE_ERROR_WARNING: /* error warning state */ priv->can.can_stats.error_warning++; priv->can.state = CAN_STATE_ERROR_WARNING; @@ -477,7 +494,7 @@ static int ifi_canfd_handle_state_change ifi_canfd_get_berr_counter(ndev, &bec); switch (new_state) { - case CAN_STATE_ERROR_ACTIVE: + case CAN_STATE_ERROR_WARNING: /* error warning state */ cf->can_id |= CAN_ERR_CRTL; cf->data[1] = (bec.txerr > bec.rxerr) ? @@ -510,22 +527,21 @@ static int ifi_canfd_handle_state_change return 1; } -static int ifi_canfd_handle_state_errors(struct net_device *ndev, u32 stcmd) +static int ifi_canfd_handle_state_errors(struct net_device *ndev) { struct ifi_canfd_priv *priv = netdev_priv(ndev); + u32 stcmd = readl(priv->base + IFI_CANFD_STCMD); int work_done = 0; - u32 isr; - /* - * The ErrWarn condition is a little special, since the bit is - * located in the INTERRUPT register instead of STCMD register. - */ - isr = readl(priv->base + IFI_CANFD_INTERRUPT); - if ((isr & IFI_CANFD_INTERRUPT_ERROR_WARNING) && + if ((stcmd & IFI_CANFD_STCMD_ERROR_ACTIVE) && + (priv->can.state != CAN_STATE_ERROR_ACTIVE)) { + netdev_dbg(ndev, "Error, entered active state\n"); + work_done += ifi_canfd_handle_state_change(ndev, + CAN_STATE_ERROR_ACTIVE); + } + + if ((stcmd & IFI_CANFD_STCMD_ERROR_WARNING) && (priv->can.state != CAN_STATE_ERROR_WARNING)) { - /* Clear the interrupt */ - writel(IFI_CANFD_INTERRUPT_ERROR_WARNING, - priv->base + IFI_CANFD_INTERRUPT); netdev_dbg(ndev, "Error, entered warning state\n"); work_done += ifi_canfd_handle_state_change(ndev, CAN_STATE_ERROR_WARNING); @@ -552,18 +568,11 @@ static int ifi_canfd_poll(struct napi_st { struct net_device *ndev = napi->dev; struct ifi_canfd_priv *priv = netdev_priv(ndev); - const u32 stcmd_state_mask = IFI_CANFD_STCMD_ERROR_PASSIVE | - IFI_CANFD_STCMD_BUSOFF; - int work_done = 0; - - u32 stcmd = readl(priv->base + IFI_CANFD_STCMD); u32 rxstcmd = readl(priv->base + IFI_CANFD_RXSTCMD); - u32 errctr = readl(priv->base + IFI_CANFD_ERROR_CTR); + int work_done = 0; /* Handle bus state changes */ - if ((stcmd & stcmd_state_mask) || - ((stcmd & IFI_CANFD_STCMD_ERROR_ACTIVE) == 0)) - work_done += ifi_canfd_handle_state_errors(ndev, stcmd); + work_done += ifi_canfd_handle_state_errors(ndev); /* Handle lost messages on RX */ if (rxstcmd & IFI_CANFD_RXSTCMD_OVERFLOW) @@ -571,7 +580,7 @@ static int ifi_canfd_poll(struct napi_st /* Handle lec errors on the bus */ if (priv->can.ctrlmode & CAN_CTRLMODE_BERR_REPORTING) - work_done += ifi_canfd_handle_lec_err(ndev, errctr); + work_done += ifi_canfd_handle_lec_err(ndev); /* Handle normal messages on RX */ if (!(rxstcmd & IFI_CANFD_RXSTCMD_EMPTY)) @@ -592,12 +601,13 @@ static irqreturn_t ifi_canfd_isr(int irq struct net_device_stats *stats = &ndev->stats; const u32 rx_irq_mask = IFI_CANFD_INTERRUPT_RXFIFO_NEMPTY | IFI_CANFD_INTERRUPT_RXFIFO_NEMPTY_PER | + IFI_CANFD_INTERRUPT_ERROR_COUNTER | + IFI_CANFD_INTERRUPT_ERROR_STATE_CHG | IFI_CANFD_INTERRUPT_ERROR_WARNING | - IFI_CANFD_INTERRUPT_ERROR_COUNTER; + IFI_CANFD_INTERRUPT_ERROR_BUSOFF; const u32 tx_irq_mask = IFI_CANFD_INTERRUPT_TXFIFO_EMPTY | IFI_CANFD_INTERRUPT_TXFIFO_REMOVE; - const u32 clr_irq_mask = ~((u32)(IFI_CANFD_INTERRUPT_SET_IRQ | - IFI_CANFD_INTERRUPT_ERROR_WARNING)); + const u32 clr_irq_mask = ~((u32)IFI_CANFD_INTERRUPT_SET_IRQ); u32 isr; isr = readl(priv->base + IFI_CANFD_INTERRUPT); Patches currently in stable-queue which might be from marex(a)denx.de are queue-4.15/can-ifi-check-core-revision-upon-probe.patch queue-4.15/can-ifi-repair-the-error-handling.patch

7 years, 9 months

1
0
0 0

Patch "can: ifi: Check core revision upon probe" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled can: ifi: Check core revision upon probe to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: can-ifi-check-core-revision-upon-probe.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 591d65d5b15496af8d05e252bc1da611c66c0b79 Mon Sep 17 00:00:00 2001 From: Marek Vasut <marex(a)denx.de> Date: Mon, 5 Mar 2018 21:29:52 +0100 Subject: can: ifi: Check core revision upon probe From: Marek Vasut <marex(a)denx.de> commit 591d65d5b15496af8d05e252bc1da611c66c0b79 upstream. Older versions of the core are not compatible with the driver due to various intrusive fixes of the core. Read out the VER register, check the core revision bitfield and verify if the core in use is new enough (rev 2.1 or newer) to work correctly with this driver. Signed-off-by: Marek Vasut <marex(a)denx.de> Cc: Heiko Schocher <hs(a)denx.de> Cc: Markus Marb <markus(a)marb.org> Cc: Marc Kleine-Budde <mkl(a)pengutronix.de> Cc: linux-stable <stable(a)vger.kernel.org> Signed-off-by: Marc Kleine-Budde <mkl(a)pengutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/can/ifi_canfd/ifi_canfd.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) --- a/drivers/net/can/ifi_canfd/ifi_canfd.c +++ b/drivers/net/can/ifi_canfd/ifi_canfd.c @@ -144,6 +144,8 @@ #define IFI_CANFD_SYSCLOCK 0x50 #define IFI_CANFD_VER 0x54 +#define IFI_CANFD_VER_REV_MASK 0xff +#define IFI_CANFD_VER_REV_MIN_SUPPORTED 0x15 #define IFI_CANFD_IP_ID 0x58 #define IFI_CANFD_IP_ID_VALUE 0xD073CAFD @@ -943,7 +945,7 @@ static int ifi_canfd_plat_probe(struct p struct resource *res; void __iomem *addr; int irq, ret; - u32 id; + u32 id, rev; res = platform_get_resource(pdev, IORESOURCE_MEM, 0); addr = devm_ioremap_resource(dev, res); @@ -957,6 +959,13 @@ static int ifi_canfd_plat_probe(struct p return -EINVAL; } + rev = readl(addr + IFI_CANFD_VER) & IFI_CANFD_VER_REV_MASK; + if (rev < IFI_CANFD_VER_REV_MIN_SUPPORTED) { + dev_err(dev, "This block is too old (rev %i), minimum supported is rev %i\n", + rev, IFI_CANFD_VER_REV_MIN_SUPPORTED); + return -EINVAL; + } + ndev = alloc_candev(sizeof(*priv), 1); if (!ndev) return -ENOMEM; Patches currently in stable-queue which might be from marex(a)denx.de are queue-4.15/can-ifi-check-core-revision-upon-probe.patch queue-4.15/can-ifi-repair-the-error-handling.patch

7 years, 9 months

1
0
0 0

Patch "can: cc770: Fix use after free in cc770_tx_interrupt()" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled can: cc770: Fix use after free in cc770_tx_interrupt() to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: can-cc770-fix-use-after-free-in-cc770_tx_interrupt.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 9ffd7503944ec7c0ef41c3245d1306c221aef2be Mon Sep 17 00:00:00 2001 From: Andri Yngvason <andri.yngvason(a)marel.com> Date: Thu, 15 Mar 2018 18:23:17 +0000 Subject: can: cc770: Fix use after free in cc770_tx_interrupt() From: Andri Yngvason <andri.yngvason(a)marel.com> commit 9ffd7503944ec7c0ef41c3245d1306c221aef2be upstream. This fixes use after free introduced by the last cc770 patch. Signed-off-by: Andri Yngvason <andri.yngvason(a)marel.com> Fixes: 746201235b3f ("can: cc770: Fix queue stall & dropped RTR reply") Cc: linux-stable <stable(a)vger.kernel.org> Signed-off-by: Marc Kleine-Budde <mkl(a)pengutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/can/cc770/cc770.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) --- a/drivers/net/can/cc770/cc770.c +++ b/drivers/net/can/cc770/cc770.c @@ -706,13 +706,12 @@ static void cc770_tx_interrupt(struct ne return; } - can_put_echo_skb(priv->tx_skb, dev, 0); - can_get_echo_skb(dev, 0); - cf = (struct can_frame *)priv->tx_skb->data; stats->tx_bytes += cf->can_dlc; stats->tx_packets++; + can_put_echo_skb(priv->tx_skb, dev, 0); + can_get_echo_skb(dev, 0); priv->tx_skb = NULL; netif_wake_queue(dev); Patches currently in stable-queue which might be from andri.yngvason(a)marel.com are queue-4.15/can-cc770-fix-use-after-free-in-cc770_tx_interrupt.patch queue-4.15/can-cc770-fix-queue-stall-dropped-rtr-reply.patch queue-4.15/can-cc770-fix-stalls-on-rt-linux-remove-redundant-irq-ack.patch

7 years, 9 months

1
0
0 0

Patch "can: cc770: Fix stalls on rt-linux, remove redundant IRQ ack" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled can: cc770: Fix stalls on rt-linux, remove redundant IRQ ack to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: can-cc770-fix-stalls-on-rt-linux-remove-redundant-irq-ack.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From f4353daf4905c0099fd25fa742e2ffd4a4bab26a Mon Sep 17 00:00:00 2001 From: Andri Yngvason <andri.yngvason(a)marel.com> Date: Wed, 14 Mar 2018 11:52:56 +0000 Subject: can: cc770: Fix stalls on rt-linux, remove redundant IRQ ack From: Andri Yngvason <andri.yngvason(a)marel.com> commit f4353daf4905c0099fd25fa742e2ffd4a4bab26a upstream. This has been reported to cause stalls on rt-linux. Suggested-by: Richard Weinberger <richard(a)nod.at> Tested-by: Richard Weinberger <richard(a)nod.at> Signed-off-by: Andri Yngvason <andri.yngvason(a)marel.com> Cc: linux-stable <stable(a)vger.kernel.org> Signed-off-by: Marc Kleine-Budde <mkl(a)pengutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/can/cc770/cc770.c | 15 --------------- 1 file changed, 15 deletions(-) --- a/drivers/net/can/cc770/cc770.c +++ b/drivers/net/can/cc770/cc770.c @@ -447,15 +447,6 @@ static netdev_tx_t cc770_start_xmit(stru stats->tx_bytes += dlc; - - /* - * HM: We had some cases of repeated IRQs so make sure the - * INT is acknowledged I know it's already further up, but - * doing again fixed the issue - */ - cc770_write_reg(priv, msgobj[mo].ctrl0, - MSGVAL_UNC | TXIE_UNC | RXIE_UNC | INTPND_RES); - return NETDEV_TX_OK; } @@ -684,12 +675,6 @@ static void cc770_tx_interrupt(struct ne /* Nothing more to send, switch off interrupts */ cc770_write_reg(priv, msgobj[mo].ctrl0, MSGVAL_RES | TXIE_RES | RXIE_RES | INTPND_RES); - /* - * We had some cases of repeated IRQ so make sure the - * INT is acknowledged - */ - cc770_write_reg(priv, msgobj[mo].ctrl0, - MSGVAL_UNC | TXIE_UNC | RXIE_UNC | INTPND_RES); stats->tx_packets++; can_get_echo_skb(dev, 0); Patches currently in stable-queue which might be from andri.yngvason(a)marel.com are queue-4.15/can-cc770-fix-use-after-free-in-cc770_tx_interrupt.patch queue-4.15/can-cc770-fix-queue-stall-dropped-rtr-reply.patch queue-4.15/can-cc770-fix-stalls-on-rt-linux-remove-redundant-irq-ack.patch

7 years, 9 months

1
0
0 0

Patch "can: cc770: Fix queue stall & dropped RTR reply" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled can: cc770: Fix queue stall & dropped RTR reply to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: can-cc770-fix-queue-stall-dropped-rtr-reply.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 746201235b3f876792099079f4c6fea941d76183 Mon Sep 17 00:00:00 2001 From: Andri Yngvason <andri.yngvason(a)marel.com> Date: Wed, 14 Mar 2018 11:52:57 +0000 Subject: can: cc770: Fix queue stall & dropped RTR reply From: Andri Yngvason <andri.yngvason(a)marel.com> commit 746201235b3f876792099079f4c6fea941d76183 upstream. While waiting for the TX object to send an RTR, an external message with a matching id can overwrite the TX data. In this case we must call the rx routine and then try transmitting the message that was overwritten again. The queue was being stalled because the RX event did not generate an interrupt to wake up the queue again and the TX event did not happen because the TXRQST flag is reset by the chip when new data is received. According to the CC770 datasheet the id of a message object should not be changed while the MSGVAL bit is set. This has been fixed by resetting the MSGVAL bit before modifying the object in the transmit function and setting it after. It is not enough to set & reset CPUUPD. It is important to keep the MSGVAL bit reset while the message object is being modified. Otherwise, during RTR transmission, a frame with matching id could trigger an rx-interrupt, which would cause a race condition between the interrupt routine and the transmit function. Signed-off-by: Andri Yngvason <andri.yngvason(a)marel.com> Tested-by: Richard Weinberger <richard(a)nod.at> Cc: linux-stable <stable(a)vger.kernel.org> Signed-off-by: Marc Kleine-Budde <mkl(a)pengutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/can/cc770/cc770.c | 94 +++++++++++++++++++++++++++++------------- drivers/net/can/cc770/cc770.h | 2 2 files changed, 68 insertions(+), 28 deletions(-) --- a/drivers/net/can/cc770/cc770.c +++ b/drivers/net/can/cc770/cc770.c @@ -390,37 +390,23 @@ static int cc770_get_berr_counter(const return 0; } -static netdev_tx_t cc770_start_xmit(struct sk_buff *skb, struct net_device *dev) +static void cc770_tx(struct net_device *dev, int mo) { struct cc770_priv *priv = netdev_priv(dev); - struct net_device_stats *stats = &dev->stats; - struct can_frame *cf = (struct can_frame *)skb->data; - unsigned int mo = obj2msgobj(CC770_OBJ_TX); + struct can_frame *cf = (struct can_frame *)priv->tx_skb->data; u8 dlc, rtr; u32 id; int i; - if (can_dropped_invalid_skb(dev, skb)) - return NETDEV_TX_OK; - - if ((cc770_read_reg(priv, - msgobj[mo].ctrl1) & TXRQST_UNC) == TXRQST_SET) { - netdev_err(dev, "TX register is still occupied!\n"); - return NETDEV_TX_BUSY; - } - - netif_stop_queue(dev); - dlc = cf->can_dlc; id = cf->can_id; - if (cf->can_id & CAN_RTR_FLAG) - rtr = 0; - else - rtr = MSGCFG_DIR; + rtr = cf->can_id & CAN_RTR_FLAG ? 0 : MSGCFG_DIR; + + cc770_write_reg(priv, msgobj[mo].ctrl0, + MSGVAL_RES | TXIE_RES | RXIE_RES | INTPND_RES); cc770_write_reg(priv, msgobj[mo].ctrl1, RMTPND_RES | TXRQST_RES | CPUUPD_SET | NEWDAT_RES); - cc770_write_reg(priv, msgobj[mo].ctrl0, - MSGVAL_SET | TXIE_SET | RXIE_RES | INTPND_RES); + if (id & CAN_EFF_FLAG) { id &= CAN_EFF_MASK; cc770_write_reg(priv, msgobj[mo].config, @@ -439,13 +425,30 @@ static netdev_tx_t cc770_start_xmit(stru for (i = 0; i < dlc; i++) cc770_write_reg(priv, msgobj[mo].data[i], cf->data[i]); - /* Store echo skb before starting the transfer */ - can_put_echo_skb(skb, dev, 0); - cc770_write_reg(priv, msgobj[mo].ctrl1, - RMTPND_RES | TXRQST_SET | CPUUPD_RES | NEWDAT_UNC); + RMTPND_UNC | TXRQST_SET | CPUUPD_RES | NEWDAT_UNC); + cc770_write_reg(priv, msgobj[mo].ctrl0, + MSGVAL_SET | TXIE_SET | RXIE_SET | INTPND_UNC); +} + +static netdev_tx_t cc770_start_xmit(struct sk_buff *skb, struct net_device *dev) +{ + struct cc770_priv *priv = netdev_priv(dev); + unsigned int mo = obj2msgobj(CC770_OBJ_TX); + + if (can_dropped_invalid_skb(dev, skb)) + return NETDEV_TX_OK; + + netif_stop_queue(dev); + + if ((cc770_read_reg(priv, + msgobj[mo].ctrl1) & TXRQST_UNC) == TXRQST_SET) { + netdev_err(dev, "TX register is still occupied!\n"); + return NETDEV_TX_BUSY; + } - stats->tx_bytes += dlc; + priv->tx_skb = skb; + cc770_tx(dev, mo); return NETDEV_TX_OK; } @@ -671,13 +674,47 @@ static void cc770_tx_interrupt(struct ne struct cc770_priv *priv = netdev_priv(dev); struct net_device_stats *stats = &dev->stats; unsigned int mo = obj2msgobj(o); + struct can_frame *cf; + u8 ctrl1; + + ctrl1 = cc770_read_reg(priv, msgobj[mo].ctrl1); - /* Nothing more to send, switch off interrupts */ cc770_write_reg(priv, msgobj[mo].ctrl0, MSGVAL_RES | TXIE_RES | RXIE_RES | INTPND_RES); + cc770_write_reg(priv, msgobj[mo].ctrl1, + RMTPND_RES | TXRQST_RES | MSGLST_RES | NEWDAT_RES); - stats->tx_packets++; + if (unlikely(!priv->tx_skb)) { + netdev_err(dev, "missing tx skb in tx interrupt\n"); + return; + } + + if (unlikely(ctrl1 & MSGLST_SET)) { + stats->rx_over_errors++; + stats->rx_errors++; + } + + /* When the CC770 is sending an RTR message and it receives a regular + * message that matches the id of the RTR message, it will overwrite the + * outgoing message in the TX register. When this happens we must + * process the received message and try to transmit the outgoing skb + * again. + */ + if (unlikely(ctrl1 & NEWDAT_SET)) { + cc770_rx(dev, mo, ctrl1); + cc770_tx(dev, mo); + return; + } + + can_put_echo_skb(priv->tx_skb, dev, 0); can_get_echo_skb(dev, 0); + + cf = (struct can_frame *)priv->tx_skb->data; + stats->tx_bytes += cf->can_dlc; + stats->tx_packets++; + + priv->tx_skb = NULL; + netif_wake_queue(dev); } @@ -789,6 +826,7 @@ struct net_device *alloc_cc770dev(int si priv->can.do_set_bittiming = cc770_set_bittiming; priv->can.do_set_mode = cc770_set_mode; priv->can.ctrlmode_supported = CAN_CTRLMODE_3_SAMPLES; + priv->tx_skb = NULL; memcpy(priv->obj_flags, cc770_obj_flags, sizeof(cc770_obj_flags)); --- a/drivers/net/can/cc770/cc770.h +++ b/drivers/net/can/cc770/cc770.h @@ -193,6 +193,8 @@ struct cc770_priv { u8 cpu_interface; /* CPU interface register */ u8 clkout; /* Clock out register */ u8 bus_config; /* Bus conffiguration register */ + + struct sk_buff *tx_skb; }; struct net_device *alloc_cc770dev(int sizeof_priv); Patches currently in stable-queue which might be from andri.yngvason(a)marel.com are queue-4.15/can-cc770-fix-use-after-free-in-cc770_tx_interrupt.patch queue-4.15/can-cc770-fix-queue-stall-dropped-rtr-reply.patch queue-4.15/can-cc770-fix-stalls-on-rt-linux-remove-redundant-irq-ack.patch

7 years, 9 months

1
0
0 0

Patch "brcmfmac: fix P2P_DEVICE ethernet address generation" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled brcmfmac: fix P2P_DEVICE ethernet address generation to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: brcmfmac-fix-p2p_device-ethernet-address-generation.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 455f3e76cfc0d893585a5f358b9ddbe9c1e1e53b Mon Sep 17 00:00:00 2001 From: Arend Van Spriel <arend.vanspriel(a)broadcom.com> Date: Wed, 28 Feb 2018 21:15:20 +0100 Subject: brcmfmac: fix P2P_DEVICE ethernet address generation From: Arend Van Spriel <arend.vanspriel(a)broadcom.com> commit 455f3e76cfc0d893585a5f358b9ddbe9c1e1e53b upstream. The firmware has a requirement that the P2P_DEVICE address should be different from the address of the primary interface. When not specified by user-space, the driver generates the MAC address for the P2P_DEVICE interface using the MAC address of the primary interface and setting the locally administered bit. However, the MAC address of the primary interface may already have that bit set causing the creation of the P2P_DEVICE interface to fail with -EBUSY. Fix this by using a random address instead to determine the P2P_DEVICE address. Cc: stable(a)vger.kernel.org # 3.10.y Reported-by: Hans de Goede <hdegoede(a)redhat.com> Reviewed-by: Hante Meuleman <hante.meuleman(a)broadcom.com> Reviewed-by: Pieter-Paul Giesberts <pieter-paul.giesberts(a)broadcom.com> Reviewed-by: Franky Lin <franky.lin(a)broadcom.com> Signed-off-by: Arend van Spriel <arend.vanspriel(a)broadcom.com> Signed-off-by: Kalle Valo <kvalo(a)codeaurora.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/wireless/broadcom/brcm80211/brcmfmac/p2p.c | 24 +++++++---------- 1 file changed, 11 insertions(+), 13 deletions(-) --- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/p2p.c +++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/p2p.c @@ -462,25 +462,23 @@ static int brcmf_p2p_set_firmware(struct * @dev_addr: optional device address. * * P2P needs mac addresses for P2P device and interface. If no device - * address it specified, these are derived from the primary net device, ie. - * the permanent ethernet address of the device. + * address it specified, these are derived from a random ethernet + * address. */ static void brcmf_p2p_generate_bss_mac(struct brcmf_p2p_info *p2p, u8 *dev_addr) { - struct brcmf_if *pri_ifp = p2p->bss_idx[P2PAPI_BSSCFG_PRIMARY].vif->ifp; - bool local_admin = false; + bool random_addr = false; - if (!dev_addr || is_zero_ether_addr(dev_addr)) { - dev_addr = pri_ifp->mac_addr; - local_admin = true; - } + if (!dev_addr || is_zero_ether_addr(dev_addr)) + random_addr = true; - /* Generate the P2P Device Address. This consists of the device's - * primary MAC address with the locally administered bit set. + /* Generate the P2P Device Address obtaining a random ethernet + * address with the locally administered bit set. */ - memcpy(p2p->dev_addr, dev_addr, ETH_ALEN); - if (local_admin) - p2p->dev_addr[0] |= 0x02; + if (random_addr) + eth_random_addr(p2p->dev_addr); + else + memcpy(p2p->dev_addr, dev_addr, ETH_ALEN); /* Generate the P2P Interface Address. If the discovery and connection * BSSCFGs need to simultaneously co-exist, then this address must be Patches currently in stable-queue which might be from arend.vanspriel(a)broadcom.com are queue-4.15/brcmfmac-fix-p2p_device-ethernet-address-generation.patch

7 years, 9 months

1
0
0 0

Patch "ACPI / watchdog: Fix off-by-one error at resource assignment" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ACPI / watchdog: Fix off-by-one error at resource assignment to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: acpi-watchdog-fix-off-by-one-error-at-resource-assignment.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From b1abf6fc49829d89660c961fafe3f90f3d843c55 Mon Sep 17 00:00:00 2001 From: Takashi Iwai <tiwai(a)suse.de> Date: Mon, 19 Mar 2018 14:51:49 +0100 Subject: ACPI / watchdog: Fix off-by-one error at resource assignment From: Takashi Iwai <tiwai(a)suse.de> commit b1abf6fc49829d89660c961fafe3f90f3d843c55 upstream. The resource allocation in WDAT watchdog has off-one-by error, it sets one byte more than the actual end address. This may eventually lead to unexpected resource conflicts. Fixes: 058dfc767008 (ACPI / watchdog: Add support for WDAT hardware watchdog) Cc: 4.9+ <stable(a)vger.kernel.org> # 4.9+ Signed-off-by: Takashi Iwai <tiwai(a)suse.de> Acked-by: Mika Westerberg <mika.westerberg(a)linux.intel.com> Acked-by: Guenter Roeck <linux(a)roeck-us.net> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki(a)intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/acpi/acpi_watchdog.c | 4 ++-- drivers/watchdog/wdat_wdt.c | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) --- a/drivers/acpi/acpi_watchdog.c +++ b/drivers/acpi/acpi_watchdog.c @@ -74,10 +74,10 @@ void __init acpi_watchdog_init(void) res.start = gas->address; if (gas->space_id == ACPI_ADR_SPACE_SYSTEM_MEMORY) { res.flags = IORESOURCE_MEM; - res.end = res.start + ALIGN(gas->access_width, 4); + res.end = res.start + ALIGN(gas->access_width, 4) - 1; } else if (gas->space_id == ACPI_ADR_SPACE_SYSTEM_IO) { res.flags = IORESOURCE_IO; - res.end = res.start + gas->access_width; + res.end = res.start + gas->access_width - 1; } else { pr_warn("Unsupported address space: %u\n", gas->space_id); --- a/drivers/watchdog/wdat_wdt.c +++ b/drivers/watchdog/wdat_wdt.c @@ -392,7 +392,7 @@ static int wdat_wdt_probe(struct platfor memset(&r, 0, sizeof(r)); r.start = gas->address; - r.end = r.start + gas->access_width; + r.end = r.start + gas->access_width - 1; if (gas->space_id == ACPI_ADR_SPACE_SYSTEM_MEMORY) { r.flags = IORESOURCE_MEM; } else if (gas->space_id == ACPI_ADR_SPACE_SYSTEM_IO) { Patches currently in stable-queue which might be from tiwai(a)suse.de are queue-4.15/alsa-aloop-fix-access-to-not-yet-ready-substream-via-cable.patch queue-4.15/alsa-hda-realtek-fix-speaker-no-sound-after-system-resume.patch queue-4.15/bluetooth-btusb-fix-quirk-for-atheros-1525-qca6174.patch queue-4.15/alsa-hda-realtek-fix-dell-headset-mic-can-t-record.patch queue-4.15/alsa-hda-realtek-always-immediately-update-mute-led-with-pin-vref.patch queue-4.15/alsa-usb-audio-fix-parsing-descriptor-of-uac2-processing-unit.patch queue-4.15/alsa-aloop-sync-stale-timer-before-release.patch queue-4.15/acpi-watchdog-fix-off-by-one-error-at-resource-assignment.patch queue-4.15/alsa-hda-force-polling-mode-on-cfl-for-fixing-codec-communication.patch

7 years, 9 months

1
0
0 0

Patch "acpi, numa: fix pxm to online numa node associations" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled acpi, numa: fix pxm to online numa node associations to the 4.15-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: acpi-numa-fix-pxm-to-online-numa-node-associations.patch and it can be found in the queue-4.15 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From dc9e0a9347e932e3fd3cd03e7ff241022ed6ea8a Mon Sep 17 00:00:00 2001 From: Dan Williams <dan.j.williams(a)intel.com> Date: Thu, 15 Mar 2018 19:49:14 -0700 Subject: acpi, numa: fix pxm to online numa node associations From: Dan Williams <dan.j.williams(a)intel.com> commit dc9e0a9347e932e3fd3cd03e7ff241022ed6ea8a upstream. Commit 99759869faf1 "acpi: Add acpi_map_pxm_to_online_node()" added support for mapping a given proximity to its nearest, by SLIT distance, online node. However, it sometimes returns unexpected results due to the fact that it switches from comparing the PXM node to the last node that was closer than the current max. for_each_online_node(n) { dist = node_distance(node, n); if (dist < min_dist) { min_dist = dist; node = n; <---- from this point we're using the wrong node for node_distance() Fixes: 99759869faf1 ("acpi: Add acpi_map_pxm_to_online_node()") Cc: <stable(a)vger.kernel.org> Reviewed-by: Toshi Kani <toshi.kani(a)hp.com> Acked-by: Rafael J. Wysocki <rafael.j.wysocki(a)intel.com>> Signed-off-by: Dan Williams <dan.j.williams(a)intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/acpi/numa.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) --- a/drivers/acpi/numa.c +++ b/drivers/acpi/numa.c @@ -103,25 +103,27 @@ int acpi_map_pxm_to_node(int pxm) */ int acpi_map_pxm_to_online_node(int pxm) { - int node, n, dist, min_dist; + int node, min_node; node = acpi_map_pxm_to_node(pxm); if (node == NUMA_NO_NODE) node = 0; + min_node = node; if (!node_online(node)) { - min_dist = INT_MAX; + int min_dist = INT_MAX, dist, n; + for_each_online_node(n) { dist = node_distance(node, n); if (dist < min_dist) { min_dist = dist; - node = n; + min_node = n; } } } - return node; + return min_node; } EXPORT_SYMBOL(acpi_map_pxm_to_online_node); Patches currently in stable-queue which might be from dan.j.williams(a)intel.com are queue-4.15/acpi-numa-fix-pxm-to-online-numa-node-associations.patch queue-4.15/libnvdimm-btt-blk-do-integrity-setup-before-add_disk.patch

7 years, 9 months

1
0
0 0

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror