September 2018 - Linux-stable-mirror

patch "USB: add quirk for WORLDE Controller KS49 or Prodipe MIDI 49C USB" added to usb-linus

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled USB: add quirk for WORLDE Controller KS49 or Prodipe MIDI 49C USB to my usb git tree which can be found at git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb.git in the usb-linus branch. The patch will show up in the next release of the linux-next tree (usually sometime within the next 24 hours during the week.) The patch will hopefully also be merged in Linus's tree for the next -rc kernel release. If you have any questions about this process, please let me know. >From 9b83a1c301ad6d24988a128c69b42cbaaf537d82 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Maxence=20Dupr=C3=A8s?= <xpros64(a)hotmail.fr> Date: Wed, 8 Aug 2018 23:56:33 +0000 Subject: USB: add quirk for WORLDE Controller KS49 or Prodipe MIDI 49C USB controller WORLDE Controller KS49 or Prodipe MIDI 49C USB controller cause a -EPROTO error, a communication restart and loop again. This issue has already been fixed for KS25. https://lore.kernel.org/patchwork/patch/753077/ I just add device 201 for KS49 in quirks.c to get it works. Signed-off-by: Laurent Roux <xpros64(a)hotmail.fr> Cc: stable <stable(a)vger.kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/usb/core/quirks.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/usb/core/quirks.c b/drivers/usb/core/quirks.c index 097057d2eacf..689a6c65bc5c 100644 --- a/drivers/usb/core/quirks.c +++ b/drivers/usb/core/quirks.c @@ -178,6 +178,10 @@ static const struct usb_device_id usb_quirk_list[] = { /* CBM - Flash disk */ { USB_DEVICE(0x0204, 0x6025), .driver_info = USB_QUIRK_RESET_RESUME }, + /* WORLDE Controller KS49 or Prodipe MIDI 49C USB controller */ + { USB_DEVICE(0x0218, 0x0201), .driver_info = + USB_QUIRK_CONFIG_INTF_STRINGS }, + /* WORLDE easy key (easykey.25) MIDI controller */ { USB_DEVICE(0x0218, 0x0401), .driver_info = USB_QUIRK_CONFIG_INTF_STRINGS }, -- 2.18.0

5 years, 8 months

1
0
0 0

[PATCH] drm/i915/bdw: Increase IPS disable timeout to 100ms

by Imre Deak

During IPS disabling the current 42ms timeout value leads to occasional timeouts, increase it to 100ms which seems to get rid of the problem. References: https://bugs.freedesktop.org/show_bug.cgi?id=107494 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107562 Reported-by: Diego Viola <diego.viola(a)gmail.com> Tested-by: Diego Viola <diego.viola(a)gmail.com> Cc: Diego Viola <diego.viola(a)gmail.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Imre Deak <imre.deak(a)intel.com> --- drivers/gpu/drm/i915/intel_display.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index a85a83f31979..1bd14c61dab5 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -5082,10 +5082,14 @@ void hsw_disable_ips(const struct intel_crtc_state *crtc_state) mutex_lock(&dev_priv->pcu_lock); WARN_ON(sandybridge_pcode_write(dev_priv, DISPLAY_IPS_CONTROL, 0)); mutex_unlock(&dev_priv->pcu_lock); - /* wait for pcode to finish disabling IPS, which may take up to 42ms */ + /* + * Wait for PCODE to finish disabling IPS. The BSpec specified + * 42ms timeout value leads to occasional timeouts so use 100ms + * instead. + */ if (intel_wait_for_register(dev_priv, IPS_CTL, IPS_ENABLE, 0, - 42)) + 100)) DRM_ERROR("Timed out waiting for IPS disable\n"); } else { I915_WRITE(IPS_CTL, 0); -- 2.13.2

5 years, 8 months

2
1
0 0

Re: [PATCH v4.9.y] Fixes: Commit 3c226c637b69 ("mm: numa: avoid waiting on freed migrated pages")

by Greg Kroah-Hartman

On Wed, Sep 05, 2018 at 05:23:28AM -0400, Chas Williams wrote: > On Wed, Sep 5, 2018 at 5:18 AM Nathan Chancellor <natechancellor(a)gmail.com> > wrote: > > > On Wed, Sep 05, 2018 at 11:05:15AM +0200, Greg KH wrote: > > > On Wed, Sep 05, 2018 at 04:58:52AM -0400, Chas Williams wrote: > > > > From: Chas Williams <chas3(a)att.com> > > > > > > > > Commit 3c226c637b69 ("mm: numa: avoid waiting on freed migrated pages") > > > > was an incomplete backport of the upstream commit. It is necessary to > > > > always reset page_nid before attempting any early exit. > > > > --- > > > > mm/huge_memory.c | 2 +- > > > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > > > > > > > <formletter> > > > > > > This is not the correct way to submit patches for inclusion in the > > > stable kernel tree. Please read: > > > > > https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html > > > for how to do this properly. > > > > > > </formletter> > > > > This is an issue with the 4.9 tree, not mainline. The hash is > > incorrect but the problem appears valid. > > > > Unfortunately, this issue doesn't fit into option 1, 2 or 3. I could > only do option 4. Ok, but if so, you need to at least sign-off on your patch, right? :) Also, call it out explicitly what you are trying to do here, and why the patch is not upstream, so I know what is going on. thanks, greg k-h

5 years, 8 months

2
2
0 0

[PATCH V4 09/10] MIPS: Loongson-3: Fix BRIDGE irq delivery problem

by Huacai Chen

After commit e509bd7da149dc349160 ("genirq: Allow migration of chained interrupts by installing default action") Loongson-3 fails at here: setup_irq(LOONGSON_HT1_IRQ, &cascade_irqaction); This is because both chained_action and cascade_irqaction don't have IRQF_SHARED flag. This will cause Loongson-3 resume fails because HPET timer interrupt can't be delivered during S3. So we set the irqchip of the chained irq to loongson_irq_chip which doesn't disable the chained irq in CP0.Status. Cc: stable(a)vger.kernel.org Signed-off-by: Huacai Chen <chenhc(a)lemote.com> --- arch/mips/include/asm/mach-loongson64/irq.h | 2 +- arch/mips/loongson64/loongson-3/irq.c | 13 +++---------- 2 files changed, 4 insertions(+), 11 deletions(-) diff --git a/arch/mips/include/asm/mach-loongson64/irq.h b/arch/mips/include/asm/mach-loongson64/irq.h index 3644b68..be9f727 100644 --- a/arch/mips/include/asm/mach-loongson64/irq.h +++ b/arch/mips/include/asm/mach-loongson64/irq.h @@ -10,7 +10,7 @@ #define MIPS_CPU_IRQ_BASE 56 #define LOONGSON_UART_IRQ (MIPS_CPU_IRQ_BASE + 2) /* UART */ -#define LOONGSON_HT1_IRQ (MIPS_CPU_IRQ_BASE + 3) /* HT1 */ +#define LOONGSON_BRIDGE_IRQ (MIPS_CPU_IRQ_BASE + 3) /* CASCADE */ #define LOONGSON_TIMER_IRQ (MIPS_CPU_IRQ_BASE + 7) /* CPU Timer */ #define LOONGSON_HT1_CFG_BASE loongson_sysconf.ht_control_base diff --git a/arch/mips/loongson64/loongson-3/irq.c b/arch/mips/loongson64/loongson-3/irq.c index 2e115ab..5605061 100644 --- a/arch/mips/loongson64/loongson-3/irq.c +++ b/arch/mips/loongson64/loongson-3/irq.c @@ -96,12 +96,6 @@ void mach_irq_dispatch(unsigned int pending) } } -static struct irqaction cascade_irqaction = { - .handler = no_action, - .flags = IRQF_NO_SUSPEND, - .name = "cascade", -}; - static inline void mask_loongson_irq(struct irq_data *d) { } static inline void unmask_loongson_irq(struct irq_data *d) { } @@ -147,11 +141,10 @@ void __init mach_init_irq(void) irq_set_chip_and_handler(LOONGSON_UART_IRQ, &loongson_irq_chip, handle_percpu_irq); + irq_set_chip_and_handler(LOONGSON_BRIDGE_IRQ, + &loongson_irq_chip, handle_percpu_irq); - /* setup HT1 irq */ - setup_irq(LOONGSON_HT1_IRQ, &cascade_irqaction); - - set_c0_status(STATUSF_IP2 | STATUSF_IP6); + set_c0_status(STATUSF_IP2 | STATUSF_IP3 | STATUSF_IP6); } #ifdef CONFIG_HOTPLUG_CPU -- 2.7.0

5 years, 8 months

1
0
0 0

[PATCH V4 08/10] MIPS: Loongson-3: Fix CPU UART irq delivery problem

by Huacai Chen

Masking/unmasking the CPU UART irq in CP0_Status (and redirecting it to other CPUs) may cause interrupts be lost, especially in multi-package machines (Package-0's UART irq cannot be delivered to others). So make mask_loongson_irq() and unmask_loongson_irq() be no-ops. The original problem (UART IRQ may deliver to any core) is also because of masking/unmasking the CPU UART irq in CP0_Status. So it is safe to remove all of the stuff. Cc: stable(a)vger.kernel.org Signed-off-by: Huacai Chen <chenhc(a)lemote.com> --- arch/mips/loongson64/loongson-3/irq.c | 43 +++-------------------------------- 1 file changed, 3 insertions(+), 40 deletions(-) diff --git a/arch/mips/loongson64/loongson-3/irq.c b/arch/mips/loongson64/loongson-3/irq.c index cbeb20f..2e115ab 100644 --- a/arch/mips/loongson64/loongson-3/irq.c +++ b/arch/mips/loongson64/loongson-3/irq.c @@ -102,45 +102,8 @@ static struct irqaction cascade_irqaction = { .name = "cascade", }; -static inline void mask_loongson_irq(struct irq_data *d) -{ - clear_c0_status(0x100 << (d->irq - MIPS_CPU_IRQ_BASE)); - irq_disable_hazard(); - - /* Workaround: UART IRQ may deliver to any core */ - if (d->irq == LOONGSON_UART_IRQ) { - int cpu = smp_processor_id(); - int node_id = cpu_logical_map(cpu) / loongson_sysconf.cores_per_node; - int core_id = cpu_logical_map(cpu) % loongson_sysconf.cores_per_node; - u64 intenclr_addr = smp_group[node_id] | - (u64)(&LOONGSON_INT_ROUTER_INTENCLR); - u64 introuter_lpc_addr = smp_group[node_id] | - (u64)(&LOONGSON_INT_ROUTER_LPC); - - *(volatile u32 *)intenclr_addr = 1 << 10; - *(volatile u8 *)introuter_lpc_addr = 0x10 + (1<<core_id); - } -} - -static inline void unmask_loongson_irq(struct irq_data *d) -{ - /* Workaround: UART IRQ may deliver to any core */ - if (d->irq == LOONGSON_UART_IRQ) { - int cpu = smp_processor_id(); - int node_id = cpu_logical_map(cpu) / loongson_sysconf.cores_per_node; - int core_id = cpu_logical_map(cpu) % loongson_sysconf.cores_per_node; - u64 intenset_addr = smp_group[node_id] | - (u64)(&LOONGSON_INT_ROUTER_INTENSET); - u64 introuter_lpc_addr = smp_group[node_id] | - (u64)(&LOONGSON_INT_ROUTER_LPC); - - *(volatile u32 *)intenset_addr = 1 << 10; - *(volatile u8 *)introuter_lpc_addr = 0x10 + (1<<core_id); - } - - set_c0_status(0x100 << (d->irq - MIPS_CPU_IRQ_BASE)); - irq_enable_hazard(); -} +static inline void mask_loongson_irq(struct irq_data *d) { } +static inline void unmask_loongson_irq(struct irq_data *d) { } /* For MIPS IRQs which shared by all cores */ static struct irq_chip loongson_irq_chip = { @@ -183,7 +146,7 @@ void __init mach_init_irq(void) chip->irq_set_affinity = plat_set_irq_affinity; irq_set_chip_and_handler(LOONGSON_UART_IRQ, - &loongson_irq_chip, handle_level_irq); + &loongson_irq_chip, handle_percpu_irq); /* setup HT1 irq */ setup_irq(LOONGSON_HT1_IRQ, &cascade_irqaction); -- 2.7.0

5 years, 8 months

1
0
0 0

[PATCH V4 05/10] MIPS: Align kernel load address to 64KB

by Huacai Chen

KEXEC needs the new kernel's load address to be aligned on a page boundary (see sanity_check_segment_list()), but on MIPS the default vmlinuz load address is only explicitly aligned to 16 bytes. Since the largest PAGE_SIZE supported by MIPS kernels is 64KB, increase the alignment calculated by calc_vmlinuz_load_addr to 64KB. Cc: <stable(a)vger.kernel.org> # 2.6.36+ Signed-off-by: Huacai Chen <chenhc(a)lemote.com> --- arch/mips/boot/compressed/calc_vmlinuz_load_addr.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/arch/mips/boot/compressed/calc_vmlinuz_load_addr.c b/arch/mips/boot/compressed/calc_vmlinuz_load_addr.c index 37fe58c..542c3ed 100644 --- a/arch/mips/boot/compressed/calc_vmlinuz_load_addr.c +++ b/arch/mips/boot/compressed/calc_vmlinuz_load_addr.c @@ -13,6 +13,7 @@ #include <stdint.h> #include <stdio.h> #include <stdlib.h> +#include "../../../../include/linux/sizes.h" int main(int argc, char *argv[]) { @@ -45,11 +46,11 @@ int main(int argc, char *argv[]) vmlinuz_load_addr = vmlinux_load_addr + vmlinux_size; /* - * Align with 16 bytes: "greater than that used for any standard data - * types by a MIPS compiler." -- See MIPS Run Linux (Second Edition). + * Align with 64KB: KEXEC needs load sections to be aligned to PAGE_SIZE, + * which may be as large as 64KB depending on the kernel configuration. */ - vmlinuz_load_addr += (16 - vmlinux_size % 16); + vmlinuz_load_addr += (SZ_64K - vmlinux_size % SZ_64K); printf("0x%llx\n", vmlinuz_load_addr); -- 2.7.0

5 years, 8 months

1
0
0 0

[PATCH V4 03/10] MIPS: Ensure pmd_present() returns false after pmd_mknotpresent()

by Huacai Chen

This patch is borrowed from ARM64 to ensure pmd_present() returns false after pmd_mknotpresent(). This is needed for THP. Cc: <stable(a)vger.kernel.org> # 3.8+ References: 5bb1cc0ff9a6 ("arm64: Ensure pmd_present() returns false after pmd_mknotpresent()") Reviewed-by: James Hogan <jhogan(a)kernel.org> Signed-off-by: Huacai Chen <chenhc(a)lemote.com> --- arch/mips/include/asm/pgtable-64.h | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/arch/mips/include/asm/pgtable-64.h b/arch/mips/include/asm/pgtable-64.h index 0036ea0..93a9dce 100644 --- a/arch/mips/include/asm/pgtable-64.h +++ b/arch/mips/include/asm/pgtable-64.h @@ -265,6 +265,11 @@ static inline int pmd_bad(pmd_t pmd) static inline int pmd_present(pmd_t pmd) { +#ifdef CONFIG_MIPS_HUGE_TLB_SUPPORT + if (unlikely(pmd_val(pmd) & _PAGE_HUGE)) + return pmd_val(pmd) & _PAGE_PRESENT; +#endif + return pmd_val(pmd) != (unsigned long) invalid_pte_table; } -- 2.7.0

5 years, 8 months

1
0
0 0

FAILED: patch "[PATCH] btrfs: use correct compare function of dirty_metadata_bytes" failed to apply to 4.9-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.9-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From d814a49198eafa6163698bdd93961302f3a877a4 Mon Sep 17 00:00:00 2001 From: Ethan Lien <ethanlien(a)synology.com> Date: Mon, 2 Jul 2018 15:44:58 +0800 Subject: [PATCH] btrfs: use correct compare function of dirty_metadata_bytes We use customized, nodesize batch value to update dirty_metadata_bytes. We should also use batch version of compare function or we will easily goto fast path and get false result from percpu_counter_compare(). Fixes: e2d845211eda ("Btrfs: use percpu counter for dirty metadata count") CC: stable(a)vger.kernel.org # 4.4+ Signed-off-by: Ethan Lien <ethanlien(a)synology.com> Reviewed-by: Nikolay Borisov <nborisov(a)suse.com> Signed-off-by: David Sterba <dsterba(a)suse.com> diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 6023eed3e805..e3858b2fe014 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -959,8 +959,9 @@ static int btree_writepages(struct address_space *mapping, fs_info = BTRFS_I(mapping->host)->root->fs_info; /* this is a bit racy, but that's ok */ - ret = percpu_counter_compare(&fs_info->dirty_metadata_bytes, - BTRFS_DIRTY_METADATA_THRESH); + ret = __percpu_counter_compare(&fs_info->dirty_metadata_bytes, + BTRFS_DIRTY_METADATA_THRESH, + fs_info->dirty_metadata_batch); if (ret < 0) return 0; } @@ -4134,8 +4135,9 @@ static void __btrfs_btree_balance_dirty(struct btrfs_fs_info *fs_info, if (flush_delayed) btrfs_balance_delayed_items(fs_info); - ret = percpu_counter_compare(&fs_info->dirty_metadata_bytes, - BTRFS_DIRTY_METADATA_THRESH); + ret = __percpu_counter_compare(&fs_info->dirty_metadata_bytes, + BTRFS_DIRTY_METADATA_THRESH, + fs_info->dirty_metadata_batch); if (ret > 0) { balance_dirty_pages_ratelimited(fs_info->btree_inode->i_mapping); }

5 years, 8 months

3
2
0 0

[PATCH 1/4] HID: multitouch: fix Elan panels with 2 input modes declaration

by Benjamin Tissoires

When implementing commit 7f81c8db5489 ("HID: multitouch: simplify the settings of the various features"), I wrongly removed a test that made sure we never try to set the second InputMode feature to something else than 0. This broke badly some recent Elan panels that now forget to send the click button in some area of the touchpad. Fixes 7f81c8db5489 Link: https://bugzilla.kernel.org/show_bug.cgi?id=200899 Cc: stable(a)vger.kernel.org # v4.18+ Signed-off-by: Benjamin Tissoires <benjamin.tissoires(a)redhat.com> --- drivers/hid/hid-multitouch.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) diff --git a/drivers/hid/hid-multitouch.c b/drivers/hid/hid-multitouch.c index 40fbb7c52723..88da991ef256 100644 --- a/drivers/hid/hid-multitouch.c +++ b/drivers/hid/hid-multitouch.c @@ -1375,7 +1375,8 @@ static bool mt_need_to_apply_feature(struct hid_device *hdev, struct hid_usage *usage, enum latency_mode latency, bool surface_switch, - bool button_switch) + bool button_switch, + bool *inputmode_found) { struct mt_device *td = hid_get_drvdata(hdev); struct mt_class *cls = &td->mtclass; @@ -1387,6 +1388,14 @@ static bool mt_need_to_apply_feature(struct hid_device *hdev, switch (usage->hid) { case HID_DG_INPUTMODE: + /* + * Some elan panels wrongly declare 2 input mode features, + * and silently ignore when we set the value in the second + * field. Skip the second feature and hope for the best. + */ + if (*inputmode_found) + return false; + if (cls->quirks & MT_QUIRK_FORCE_GET_FEATURE) { report_len = hid_report_len(report); buf = hid_alloc_report_buf(report, GFP_KERNEL); @@ -1402,6 +1411,7 @@ static bool mt_need_to_apply_feature(struct hid_device *hdev, } field->value[index] = td->inputmode_value; + *inputmode_found = true; return true; case HID_DG_CONTACTMAX: @@ -1439,6 +1449,7 @@ static void mt_set_modes(struct hid_device *hdev, enum latency_mode latency, struct hid_usage *usage; int i, j; bool update_report; + bool inputmode_found = false; rep_enum = &hdev->report_enum[HID_FEATURE_REPORT]; list_for_each_entry(rep, &rep_enum->report_list, list) { @@ -1457,7 +1468,8 @@ static void mt_set_modes(struct hid_device *hdev, enum latency_mode latency, usage, latency, surface_switch, - button_switch)) + button_switch, + &inputmode_found)) update_report = true; } } -- 2.14.3

5 years, 8 months

2
2
0 0

[PATCH v6 1/2] mm: migration: fix migration of huge PMD shared pages

by Mike Kravetz

The page migration code employs try_to_unmap() to try and unmap the source page. This is accomplished by using rmap_walk to find all vmas where the page is mapped. This search stops when page mapcount is zero. For shared PMD huge pages, the page map count is always 1 no matter the number of mappings. Shared mappings are tracked via the reference count of the PMD page. Therefore, try_to_unmap stops prematurely and does not completely unmap all mappings of the source page. This problem can result is data corruption as writes to the original source page can happen after contents of the page are copied to the target page. Hence, data is lost. This problem was originally seen as DB corruption of shared global areas after a huge page was soft offlined due to ECC memory errors. DB developers noticed they could reproduce the issue by (hotplug) offlining memory used to back huge pages. A simple testcase can reproduce the problem by creating a shared PMD mapping (note that this must be at least PUD_SIZE in size and PUD_SIZE aligned (1GB on x86)), and using migrate_pages() to migrate process pages between nodes while continually writing to the huge pages being migrated. To fix, have the try_to_unmap_one routine check for huge PMD sharing by calling huge_pmd_unshare for hugetlbfs huge pages. If it is a shared mapping it will be 'unshared' which removes the page table entry and drops the reference on the PMD page. After this, flush caches and TLB. mmu notifiers are called before locking page tables, but we can not be sure of PMD sharing until page tables are locked. Therefore, check for the possibility of PMD sharing before locking so that notifiers can prepare for the worst possible case. Fixes: 39dde65c9940 ("shared page table for hugetlb page") Cc: stable(a)vger.kernel.org Signed-off-by: Mike Kravetz <mike.kravetz(a)oracle.com> --- include/linux/hugetlb.h | 14 ++++++++++++++ mm/hugetlb.c | 40 +++++++++++++++++++++++++++++++++++++-- mm/rmap.c | 42 ++++++++++++++++++++++++++++++++++++++--- 3 files changed, 91 insertions(+), 5 deletions(-) diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index 36fa6a2a82e3..4ee95d8c8413 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -140,6 +140,8 @@ pte_t *huge_pte_alloc(struct mm_struct *mm, pte_t *huge_pte_offset(struct mm_struct *mm, unsigned long addr, unsigned long sz); int huge_pmd_unshare(struct mm_struct *mm, unsigned long *addr, pte_t *ptep); +void adjust_range_if_pmd_sharing_possible(struct vm_area_struct *vma, + unsigned long *start, unsigned long *end); struct page *follow_huge_addr(struct mm_struct *mm, unsigned long address, int write); struct page *follow_huge_pd(struct vm_area_struct *vma, @@ -170,6 +172,18 @@ static inline unsigned long hugetlb_total_pages(void) return 0; } +static inline int huge_pmd_unshare(struct mm_struct *mm, unsigned long *addr, + pte_t *ptep) +{ + return 0; +} + +static inline void adjust_range_if_pmd_sharing_possible( + struct vm_area_struct *vma, + unsigned long *start, unsigned long *end) +{ +} + #define follow_hugetlb_page(m,v,p,vs,a,b,i,w,n) ({ BUG(); 0; }) #define follow_huge_addr(mm, addr, write) ERR_PTR(-EINVAL) #define copy_hugetlb_page_range(src, dst, vma) ({ BUG(); 0; }) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 3103099f64fd..a73c5728e961 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -4548,6 +4548,9 @@ static unsigned long page_table_shareable(struct vm_area_struct *svma, return saddr; } +#define _range_in_vma(vma, start, end) \ + ((vma)->vm_start <= (start) && (end) <= (vma)->vm_end) + static bool vma_shareable(struct vm_area_struct *vma, unsigned long addr) { unsigned long base = addr & PUD_MASK; @@ -4556,12 +4559,40 @@ static bool vma_shareable(struct vm_area_struct *vma, unsigned long addr) /* * check on proper vm_flags and page table alignment */ - if (vma->vm_flags & VM_MAYSHARE && - vma->vm_start <= base && end <= vma->vm_end) + if (vma->vm_flags & VM_MAYSHARE && _range_in_vma(vma, base, end)) return true; return false; } +/* + * Determine if start,end range within vma could be mapped by shared pmd. + * If yes, adjust start and end to cover range associated with possible + * shared pmd mappings. + */ +void adjust_range_if_pmd_sharing_possible(struct vm_area_struct *vma, + unsigned long *start, unsigned long *end) +{ + unsigned long check_addr = *start; + + if (!(vma->vm_flags & VM_MAYSHARE)) + return; + + for (check_addr = *start; check_addr < *end; check_addr += PUD_SIZE) { + unsigned long a_start = check_addr & PUD_MASK; + unsigned long a_end = a_start + PUD_SIZE; + + /* + * If sharing is possible, adjust start/end if necessary. + */ + if (_range_in_vma(vma, a_start, a_end)) { + if (a_start < *start) + *start = a_start; + if (a_end > *end) + *end = a_end; + } + } +} + /* * Search for a shareable pmd page for hugetlb. In any case calls pmd_alloc() * and returns the corresponding pte. While this is not necessary for the @@ -4659,6 +4690,11 @@ int huge_pmd_unshare(struct mm_struct *mm, unsigned long *addr, pte_t *ptep) { return 0; } + +void adjust_range_if_pmd_sharing_possible(struct vm_area_struct *vma, + unsigned long *start, unsigned long *end) +{ +} #define want_pmd_share() (0) #endif /* CONFIG_ARCH_WANT_HUGE_PMD_SHARE */ diff --git a/mm/rmap.c b/mm/rmap.c index eb477809a5c0..1e79fac3186b 100644 --- a/mm/rmap.c +++ b/mm/rmap.c @@ -1362,11 +1362,21 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma, } /* - * We have to assume the worse case ie pmd for invalidation. Note that - * the page can not be free in this function as call of try_to_unmap() - * must hold a reference on the page. + * For THP, we have to assume the worse case ie pmd for invalidation. + * For hugetlb, it could be much worse if we need to do pud + * invalidation in the case of pmd sharing. + * + * Note that the page can not be free in this function as call of + * try_to_unmap() must hold a reference on the page. */ end = min(vma->vm_end, start + (PAGE_SIZE << compound_order(page))); + if (PageHuge(page)) { + /* + * If sharing is possible, start and end will be adjusted + * accordingly. + */ + adjust_range_if_pmd_sharing_possible(vma, &start, &end); + } mmu_notifier_invalidate_range_start(vma->vm_mm, start, end); while (page_vma_mapped_walk(&pvmw)) { @@ -1409,6 +1419,32 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma, subpage = page - page_to_pfn(page) + pte_pfn(*pvmw.pte); address = pvmw.address; + if (PageHuge(page)) { + if (huge_pmd_unshare(mm, &address, pvmw.pte)) { + /* + * huge_pmd_unshare unmapped an entire PMD + * page. There is no way of knowing exactly + * which PMDs may be cached for this mm, so + * we must flush them all. start/end were + * already adjusted above to cover this range. + */ + flush_cache_range(vma, start, end); + flush_tlb_range(vma, start, end); + mmu_notifier_invalidate_range(mm, start, end); + + /* + * The ref count of the PMD page was dropped + * which is part of the way map counting + * is done for shared PMDs. Return 'true' + * here. When there is no other sharing, + * huge_pmd_unshare returns false and we will + * unmap the actual page and drop map count + * to zero. + */ + page_vma_mapped_walk_done(&pvmw); + break; + } + } if (IS_ENABLED(CONFIG_MIGRATION) && (flags & TTU_MIGRATION) && -- 2.17.1

5 years, 8 months

4
24
0 0

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror September 2018