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

[Linux-stable-mirror] [merged] userfaultfd-clear-the-vma-vm_userfaultfd_ctx-if-uffd_event_fork-fails.patch removed from -mm tree

by akpm＠linux-foundation.org

The patch titled Subject: userfaultfd: clear the vma->vm_userfaultfd_ctx if UFFD_EVENT_FORK fails has been removed from the -mm tree. Its filename was userfaultfd-clear-the-vma-vm_userfaultfd_ctx-if-uffd_event_fork-fails.patch This patch was dropped because it was merged into mainline or a subsystem tree ------------------------------------------------------ From: Andrea Arcangeli <aarcange(a)redhat.com> Subject: userfaultfd: clear the vma->vm_userfaultfd_ctx if UFFD_EVENT_FORK fails The previous fix 384632e67e0829 ("userfaultfd: non-cooperative: fix fork use after free") corrected the refcounting in case of UFFD_EVENT_FORK failure for the fork userfault paths. That still didn't clear the vma->vm_userfaultfd_ctx of the vmas that were set to point to the aborted new uffd ctx earlier in dup_userfaultfd. Link: http://lkml.kernel.org/r/20171223002505.593-2-aarcange@redhat.com Signed-off-by: Andrea Arcangeli <aarcange(a)redhat.com> Reported-by: syzbot <syzkaller(a)googlegroups.com> Reviewed-by: Mike Rapoport <rppt(a)linux.vnet.ibm.com> Cc: Eric Biggers <ebiggers3(a)gmail.com> Cc: Dmitry Vyukov <dvyukov(a)google.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- fs/userfaultfd.c | 20 ++++++++++++++++++-- 1 file changed, 18 insertions(+), 2 deletions(-) diff -puN fs/userfaultfd.c~userfaultfd-clear-the-vma-vm_userfaultfd_ctx-if-uffd_event_fork-fails fs/userfaultfd.c --- a/fs/userfaultfd.c~userfaultfd-clear-the-vma-vm_userfaultfd_ctx-if-uffd_event_fork-fails +++ a/fs/userfaultfd.c @@ -570,11 +570,14 @@ out: static void userfaultfd_event_wait_completion(struct userfaultfd_ctx *ctx, struct userfaultfd_wait_queue *ewq) { + struct userfaultfd_ctx *release_new_ctx; + if (WARN_ON_ONCE(current->flags & PF_EXITING)) goto out; ewq->ctx = ctx; init_waitqueue_entry(&ewq->wq, current); + release_new_ctx = NULL; spin_lock(&ctx->event_wqh.lock); /* @@ -601,8 +604,7 @@ static void userfaultfd_event_wait_compl new = (struct userfaultfd_ctx *) (unsigned long) ewq->msg.arg.reserved.reserved1; - - userfaultfd_ctx_put(new); + release_new_ctx = new; } break; } @@ -617,6 +619,20 @@ static void userfaultfd_event_wait_compl __set_current_state(TASK_RUNNING); spin_unlock(&ctx->event_wqh.lock); + if (release_new_ctx) { + struct vm_area_struct *vma; + struct mm_struct *mm = release_new_ctx->mm; + + /* the various vma->vm_userfaultfd_ctx still points to it */ + down_write(&mm->mmap_sem); + for (vma = mm->mmap; vma; vma = vma->vm_next) + if (vma->vm_userfaultfd_ctx.ctx == release_new_ctx) + vma->vm_userfaultfd_ctx = NULL_VM_UFFD_CTX; + up_write(&mm->mmap_sem); + + userfaultfd_ctx_put(release_new_ctx); + } + /* * ctx may go away after this if the userfault pseudo fd is * already released. _ Patches currently in -mm which might be from aarcange(a)redhat.com are

7 years, 3 months

1
0
0 0

[Linux-stable-mirror] [merged] mm-sparsec-wrong-allocation-for-mem_section.patch removed from -mm tree

by akpm＠linux-foundation.org

The patch titled Subject: mm/sparse.c: wrong allocation for mem_section has been removed from the -mm tree. Its filename was mm-sparsec-wrong-allocation-for-mem_section.patch This patch was dropped because it was merged into mainline or a subsystem tree ------------------------------------------------------ From: Baoquan He <bhe(a)redhat.com> Subject: mm/sparse.c: wrong allocation for mem_section In 83e3c48729 ("mm/sparsemem: Allocate mem_section at runtime for CONFIG_SPARSEMEM_EXTREME=y") mem_section is allocated at runtime to save memory. While it allocates the first dimension of array with sizeof(struct mem_section). It costs extra memory, should be sizeof(struct mem_section*). Fix it. Link: http://lkml.kernel.org/r/1513932498-20350-1-git-send-email-bhe@redhat.com Fixes: 83e3c48729 ("mm/sparsemem: Allocate mem_section at runtime for CONFIG_SPARSEMEM_EXTREME=y") Signed-off-by: Baoquan He <bhe(a)redhat.com> Tested-by: Dave Young <dyoung(a)redhat.com> Acked-by: Kirill A. Shutemov <kirill.shutemov(a)linux.intel.com> Cc: Kirill A. Shutemov <kirill.shutemov(a)linux.intel.com> Cc: Ingo Molnar <mingo(a)kernel.org> Cc: Andy Lutomirski <luto(a)amacapital.net> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Thomas Gleixner <tglx(a)linutronix.de> Cc: Atsushi Kumagai <ats-kumagai(a)wm.jp.nec.com> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- mm/sparse.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff -puN mm/sparse.c~mm-sparsec-wrong-allocation-for-mem_section mm/sparse.c --- a/mm/sparse.c~mm-sparsec-wrong-allocation-for-mem_section +++ a/mm/sparse.c @@ -211,7 +211,7 @@ void __init memory_present(int nid, unsi if (unlikely(!mem_section)) { unsigned long size, align; - size = sizeof(struct mem_section) * NR_SECTION_ROOTS; + size = sizeof(struct mem_section*) * NR_SECTION_ROOTS; align = 1 << (INTERNODE_CACHE_SHIFT); mem_section = memblock_virt_alloc(size, align); } _ Patches currently in -mm which might be from bhe(a)redhat.com are

7 years, 3 months

1
0
0 0

[Linux-stable-mirror] [merged] mm-mprotect-add-a-cond_resched-inside-change_pmd_range.patch removed from -mm tree

by akpm＠linux-foundation.org

The patch titled Subject: mm/mprotect: add a cond_resched() inside change_pmd_range() has been removed from the -mm tree. Its filename was mm-mprotect-add-a-cond_resched-inside-change_pmd_range.patch This patch was dropped because it was merged into mainline or a subsystem tree ------------------------------------------------------ From: Anshuman Khandual <khandual(a)linux.vnet.ibm.com> Subject: mm/mprotect: add a cond_resched() inside change_pmd_range() While testing on a large CPU system, detected the following RCU stall many times over the span of the workload. This problem is solved by adding a cond_resched() in the change_pmd_range() function. [ 850.962530] INFO: rcu_sched detected stalls on CPUs/tasks: [ 850.962584] 154-....: (670 ticks this GP) idle=022/140000000000000/0 softirq=2825/2825 fqs=612 [ 850.962605] (detected by 955, t=6002 jiffies, g=4486, c=4485, q=90864) [ 850.962895] Sending NMI from CPU 955 to CPUs 154: [ 850.992667] NMI backtrace for cpu 154 [ 850.993069] CPU: 154 PID: 147071 Comm: workload Not tainted 4.15.0-rc3+ #3 [ 850.993258] NIP: c0000000000b3f64 LR: c0000000000b33d4 CTR: 000000000000aa18 [ 850.993503] REGS: 00000000a4b0fb44 TRAP: 0501 Not tainted (4.15.0-rc3+) [ 850.993707] MSR: 8000000000009033 <SF,EE,ME,IR,DR,RI,LE> CR: 22422082 XER: 00000000 [ 850.994386] CFAR: 00000000006cf8f0 SOFTE: 1 GPR00: 0010000000000000 c00003ef9b1cb8c0 c0000000010cc600 0000000000000000 GPR04: 8e0000018c32b200 40017b3858fd6e00 8e0000018c32b208 40017b3858fd6e00 GPR08: 8e0000018c32b210 40017b3858fd6e00 8e0000018c32b218 40017b3858fd6e00 GPR12: ffffffffffffffff c00000000fb25100 [ 850.995976] NIP [c0000000000b3f64] plpar_hcall9+0x44/0x7c [ 850.996174] LR [c0000000000b33d4] pSeries_lpar_flush_hash_range+0x384/0x420 [ 850.996401] Call Trace: [ 850.996600] [c00003ef9b1cb8c0] [c00003fa8fff7d40] 0xc00003fa8fff7d40 (unreliable) [ 850.996959] [c00003ef9b1cba40] [c0000000000688a8] flush_hash_range+0x48/0x100 [ 850.997261] [c00003ef9b1cba90] [c000000000071b14] __flush_tlb_pending+0x44/0xd0 [ 850.997600] [c00003ef9b1cbac0] [c000000000071fa8] hpte_need_flush+0x408/0x470 [ 850.997958] [c00003ef9b1cbb30] [c0000000002c646c] change_protection_range+0xaac/0xf10 [ 850.998180] [c00003ef9b1cbcb0] [c0000000002f2510] change_prot_numa+0x30/0xb0 [ 850.998502] [c00003ef9b1cbce0] [c00000000013a950] task_numa_work+0x2d0/0x3e0 [ 850.998816] [c00003ef9b1cbda0] [c00000000011ea30] task_work_run+0x130/0x190 [ 850.999121] [c00003ef9b1cbe00] [c00000000001bcd8] do_notify_resume+0x118/0x120 [ 850.999421] [c00003ef9b1cbe30] [c00000000000b744] ret_from_except_lite+0x70/0x74 [ 850.999716] Instruction dump: [ 850.999959] 60000000 f8810028 7ca42b78 7cc53378 7ce63b78 7d074378 7d284b78 7d495378 [ 851.000575] e9410060 e9610068 e9810070 44000022 <7d806378> e9810028 f88c0000 f8ac0008 Link: http://lkml.kernel.org/r/20171214140551.5794-1-khandual@linux.vnet.ibm.com Signed-off-by: Anshuman Khandual <khandual(a)linux.vnet.ibm.com> Suggested-by: Nicholas Piggin <npiggin(a)gmail.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> --- mm/mprotect.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff -puN mm/mprotect.c~mm-mprotect-add-a-cond_resched-inside-change_pmd_range mm/mprotect.c --- a/mm/mprotect.c~mm-mprotect-add-a-cond_resched-inside-change_pmd_range +++ a/mm/mprotect.c @@ -166,7 +166,7 @@ static inline unsigned long change_pmd_r next = pmd_addr_end(addr, end); if (!is_swap_pmd(*pmd) && !pmd_trans_huge(*pmd) && !pmd_devmap(*pmd) && pmd_none_or_clear_bad(pmd)) - continue; + goto next; /* invoke the mmu notifier if the pmd is populated */ if (!mni_start) { @@ -188,7 +188,7 @@ static inline unsigned long change_pmd_r } /* huge pmd was handled */ - continue; + goto next; } } /* fall through, the trans huge pmd just split */ @@ -196,6 +196,8 @@ static inline unsigned long change_pmd_r this_pages = change_pte_range(vma, pmd, addr, next, newprot, dirty_accountable, prot_numa); pages += this_pages; +next: + cond_resched(); } while (pmd++, addr = next, addr != end); if (mni_start) _ Patches currently in -mm which might be from khandual(a)linux.vnet.ibm.com are

7 years, 3 months

1
0
0 0

[Linux-stable-mirror] [merged] acct-fix-the-acct-needcheck-check-in-check_free_space.patch removed from -mm tree

by akpm＠linux-foundation.org

The patch titled Subject: kernel/acct.c: fix the acct->needcheck check in check_free_space() has been removed from the -mm tree. Its filename was acct-fix-the-acct-needcheck-check-in-check_free_space.patch This patch was dropped because it was merged into mainline or a subsystem tree ------------------------------------------------------ From: Oleg Nesterov <oleg(a)redhat.com> Subject: kernel/acct.c: fix the acct->needcheck check in check_free_space() As Tsukada explains, the time_is_before_jiffies(acct->needcheck) check is very wrong, we need time_is_after_jiffies() to make sys_acct() work. Ignoring the overflows, the code should "goto out" if needcheck > jiffies, while currently it checks "needcheck < jiffies" and thus in the likely case check_free_space() does nothing until jiffies overflow. In particular this means that sys_acct() is simply broken, acct_on() sets acct->needcheck = jiffies and expects that check_free_space() should set acct->active = 1 after the free-space check, but this won't happen if jiffies increments in between. This was broken by commit 32dc73086015 ("get rid of timer in kern/acct.c") in 2011, then another (correct) commit 795a2f22a8ea ("acct() should honour the limits from the very beginning") made the problem more visible. Link: http://lkml.kernel.org/r/20171213133940.GA6554@redhat.com Fixes: 32dc73086015 ("get rid of timer in kern/acct.c") Reported-by: TSUKADA Koutaro <tsukada(a)ascade.co.jp> Suggested-by: TSUKADA Koutaro <tsukada(a)ascade.co.jp> Signed-off-by: Oleg Nesterov <oleg(a)redhat.com> Cc: Al Viro <viro(a)zeniv.linux.org.uk> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- kernel/acct.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff -puN kernel/acct.c~acct-fix-the-acct-needcheck-check-in-check_free_space kernel/acct.c --- a/kernel/acct.c~acct-fix-the-acct-needcheck-check-in-check_free_space +++ a/kernel/acct.c @@ -102,7 +102,7 @@ static int check_free_space(struct bsd_a { struct kstatfs sbuf; - if (time_is_before_jiffies(acct->needcheck)) + if (time_is_after_jiffies(acct->needcheck)) goto out; /* May block */ _ Patches currently in -mm which might be from oleg(a)redhat.com are

7 years, 3 months

1
0
0 0

[Linux-stable-mirror] [PATCH 2/2] mwifiex: resolve reset vs. remove()/shutdown() deadlocks

by Brian Norris

Commit b014e96d1abb ("PCI: Protect pci_error_handlers->reset_notify() usage with device_lock()") resolves races between driver reset and removal, but it introduces some new deadlock problems. If we see a timeout while we've already started suspending, removing, or shutting down the driver, we might see: (a) a worker thread, running mwifiex_pcie_work() -> mwifiex_pcie_card_reset_work() -> pci_reset_function() (b) a removal thread, running mwifiex_pcie_remove() -> mwifiex_free_adapter() -> mwifiex_unregister() -> mwifiex_cleanup_pcie() -> cancel_work_sync(&card->work) Unfortunately, mwifiex_pcie_remove() already holds the device lock that pci_reset_function() is now requesting, and so we see a deadlock. It's necessary to cancel and synchronize our outstanding work before tearing down the driver, so we can't have this work wait indefinitely for the lock. It's reasonable to only "try" to reset here, since this will mostly happen for cases where it's already difficult to reset the firmware anyway (e.g., while we're suspending or powering off the system). And if reset *really* needs to happen, we can always try again later. Fixes: b014e96d1abb ("PCI: Protect pci_error_handlers->reset_notify() usage with device_lock()") Cc: <stable(a)vger.kernel.org> Cc: Xinming Hu <huxm(a)marvell.com> Signed-off-by: Brian Norris <briannorris(a)chromium.org> --- drivers/net/wireless/marvell/mwifiex/pcie.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/marvell/mwifiex/pcie.c b/drivers/net/wireless/marvell/mwifiex/pcie.c index f666cb2ea7e0..97a6199692ab 100644 --- a/drivers/net/wireless/marvell/mwifiex/pcie.c +++ b/drivers/net/wireless/marvell/mwifiex/pcie.c @@ -2786,7 +2786,10 @@ static void mwifiex_pcie_card_reset_work(struct mwifiex_adapter *adapter) { struct pcie_service_card *card = adapter->card; - pci_reset_function(card->dev); + /* We can't afford to wait here; remove() might be waiting on us. If we + * can't grab the device lock, maybe we'll get another chance later. + */ + pci_try_reset_function(card->dev); } static void mwifiex_pcie_work(struct work_struct *work) -- 2.16.0.rc1.238.g530d649a79-goog

7 years, 3 months

1
0
0 0

[Linux-stable-mirror] [PATCH 1/2] mn10300/misalignment: Use SIGSEGV SEGV_MAPERR to report a failed user copy

by Eric W. Biederman

Setting si_code to 0 is the same a setting si_code to SI_USER which is definitely not correct. With si_code set to SI_USER si_pid and si_uid will be copied to userspace instead of si_addr. Which is very wrong. So fix this by using a sensible si_code (SEGV_MAPERR) for this failure. Cc: stable(a)vger.kernel.org Fixes: b920de1b77b7 ("mn10300: add the MN10300/AM33 architecture to the kernel") Cc: David Howells <dhowells(a)redhat.com> Cc: Masakazu Urade <urade.masakazu(a)jp.panasonic.com> Cc: Koichi Yasutake <yasutake.koichi(a)jp.panasonic.com> Signed-off-by: "Eric W. Biederman" <ebiederm(a)xmission.com> --- arch/mn10300/mm/misalignment.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/mn10300/mm/misalignment.c b/arch/mn10300/mm/misalignment.c index b39a388825ae..8ace89617c1c 100644 --- a/arch/mn10300/mm/misalignment.c +++ b/arch/mn10300/mm/misalignment.c @@ -437,7 +437,7 @@ asmlinkage void misalignment(struct pt_regs *regs, enum exception_code code) info.si_signo = SIGSEGV; info.si_errno = 0; - info.si_code = 0; + info.si_code = SEGV_MAPERR; info.si_addr = (void *) regs->pc; force_sig_info(SIGSEGV, &info, current); return; -- 2.14.1

7 years, 3 months

1
0
0 0

[Linux-stable-mirror] [PATCH v2] PCI: iproc: Fix NULL pointer dereference for BCMA

by Ray Jui

With the inbound DMA mapping supported added, the iProc PCIe driver parses DT property "dma-ranges" through call to "of_pci_dma_range_parser_init". In the case of BCMA, this results in a NULL pointer deference due to a missing of_node. Fix this by adding a guard in pcie-iproc-platform.c to only enable the inbound DMA mapping logic when DT property "dma-ranges" is present Fixes: dd9d4e7498de3 ("PCI: iproc: Add inbound DMA mapping support") Reported-by: Rafał Miłecki <rafal(a)milecki.pl> Signed-off-by: Ray Jui <ray.jui(a)broadcom.com> Tested-by: Rafał Miłecki <rafal(a)milecki.pl> cc: <stable(a)vger.kernel.org> # 4.10+ --- drivers/pci/host/pcie-iproc-platform.c | 7 +++++++ drivers/pci/host/pcie-iproc.c | 8 +++++--- drivers/pci/host/pcie-iproc.h | 2 ++ 3 files changed, 14 insertions(+), 3 deletions(-) diff --git a/drivers/pci/host/pcie-iproc-platform.c b/drivers/pci/host/pcie-iproc-platform.c index a5073a9..32228d4 100644 --- a/drivers/pci/host/pcie-iproc-platform.c +++ b/drivers/pci/host/pcie-iproc-platform.c @@ -92,6 +92,13 @@ static int iproc_pcie_pltfm_probe(struct platform_device *pdev) pcie->need_ob_cfg = true; } + /* + * DT nodes are not used by all platforms that use the iProc PCIe + * core driver. For platforms that require explict inbound mapping + * configuration, "dma-ranges" would have been present in DT + */ + pcie->need_ib_cfg = of_property_read_bool(np, "dma-ranges"); + /* PHY use is optional */ pcie->phy = devm_phy_get(dev, "pcie-phy"); if (IS_ERR(pcie->phy)) { diff --git a/drivers/pci/host/pcie-iproc.c b/drivers/pci/host/pcie-iproc.c index 935909b..7583606 100644 --- a/drivers/pci/host/pcie-iproc.c +++ b/drivers/pci/host/pcie-iproc.c @@ -1378,9 +1378,11 @@ int iproc_pcie_setup(struct iproc_pcie *pcie, struct list_head *res) } } - ret = iproc_pcie_map_dma_ranges(pcie); - if (ret && ret != -ENOENT) - goto err_power_off_phy; + if (pcie->need_ib_cfg) { + ret = iproc_pcie_map_dma_ranges(pcie); + if (ret && ret != -ENOENT) + goto err_power_off_phy; + } #ifdef CONFIG_ARM pcie->sysdata.private_data = pcie; diff --git a/drivers/pci/host/pcie-iproc.h b/drivers/pci/host/pcie-iproc.h index a6b55ce..4ac6282 100644 --- a/drivers/pci/host/pcie-iproc.h +++ b/drivers/pci/host/pcie-iproc.h @@ -74,6 +74,7 @@ struct iproc_msi; * @ob: outbound mapping related parameters * @ob_map: outbound mapping related parameters specific to the controller * + * @need_ib_cfg: indicates SW needs to configure the inbound mapping window * @ib: inbound mapping related parameters * @ib_map: outbound mapping region related parameters * @@ -101,6 +102,7 @@ struct iproc_pcie { struct iproc_pcie_ob ob; const struct iproc_pcie_ob_map *ob_map; + bool need_ib_cfg; struct iproc_pcie_ib ib; const struct iproc_pcie_ib_map *ib_map; -- 2.1.4

7 years, 3 months

2
1
0 0

[Linux-stable-mirror] [PATCH 03/11] signal/openrisc: Fix do_unaligned_access to send the proper signal

by Eric W. Biederman

While reviewing the signal sending on openrisc the do_unaligned_access function stood out because it is obviously wrong. A comment about an si_code set above when actually si_code is never set. Leading to a random si_code being sent to userspace in the event of an unaligned access. Looking further SIGBUS BUS_ADRALN is the proper pair of signal and si_code to send for an unaligned access. That is what other architectures do and what is required by posix. Given that do_unaligned_access is broken in a way that no one can be relying on it on openrisc fix the code to just do the right thing. Cc: stable(a)vger.kernel.org Fixes: 769a8a96229e ("OpenRISC: Traps") Cc: Jonas Bonn <jonas(a)southpole.se> Cc: Stefan Kristiansson <stefan.kristiansson(a)saunalahti.fi> Cc: Stafford Horne <shorne(a)gmail.com> Cc: Arnd Bergmann <arnd(a)arndb.de> Cc: openrisc(a)lists.librecores.org Signed-off-by: "Eric W. Biederman" <ebiederm(a)xmission.com> --- arch/openrisc/kernel/traps.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/arch/openrisc/kernel/traps.c b/arch/openrisc/kernel/traps.c index 4085d72fa5ae..9e38dc66c9e4 100644 --- a/arch/openrisc/kernel/traps.c +++ b/arch/openrisc/kernel/traps.c @@ -266,12 +266,12 @@ asmlinkage void do_unaligned_access(struct pt_regs *regs, unsigned long address) siginfo_t info; if (user_mode(regs)) { - /* Send a SIGSEGV */ - info.si_signo = SIGSEGV; + /* Send a SIGBUS */ + info.si_signo = SIGBUS; info.si_errno = 0; - /* info.si_code has been set above */ - info.si_addr = (void *)address; - force_sig_info(SIGSEGV, &info, current); + info.si_code = BUS_ADRALN; + info.si_addr = (void __user *)address; + force_sig_info(SIGBUS, &info, current); } else { printk("KERNEL: Unaligned Access 0x%.8lx\n", address); show_registers(regs); -- 2.14.1

7 years, 3 months

3
2
0 0

[Linux-stable-mirror] [PATCH v3] mtd: nand: Fix nand_do_read_oob() return value

by Miquel Raynal

Starting from commit 041e4575f034 ("mtd: nand: handle ECC errors in OOB"), nand_do_read_oob() (from the NAND core) did return 0 or a negative error, and the MTD layer expected it. However, the trend for the NAND layer is now to return an error or a positive number of bitflips. Deciding which status to return to the user belongs to the MTD layer. Commit e47f68587b82 ("mtd: check for max_bitflips in mtd_read_oob()") brought this logic to the mtd_read_oob() function while the return value coming from nand_do_read_oob() (called by the ->_read_oob() hook) was left unchanged. Fixes: e47f68587b82 ("mtd: check for max_bitflips in mtd_read_oob()") Cc: stable(a)vger.kernel.org Signed-off-by: Miquel Raynal <miquel.raynal(a)free-electrons.com> --- Changes since v2: - Correctly return the maximum number of bitflips, not the number of bitflips of the last chunk only. Changes since v1: - s/->ecc.read_oob() hook/->_read_oob() hook/ in the commit message - Fixed the compilation issue drivers/mtd/nand/nand_base.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/mtd/nand/nand_base.c b/drivers/mtd/nand/nand_base.c index 469220065b8b..e4e39890a975 100644 --- a/drivers/mtd/nand/nand_base.c +++ b/drivers/mtd/nand/nand_base.c @@ -3794,6 +3794,7 @@ EXPORT_SYMBOL(nand_write_oob_syndrome); static int nand_do_read_oob(struct mtd_info *mtd, loff_t from, struct mtd_oob_ops *ops) { + unsigned int max_bitflips = 0; int page, realpage, chipnr; struct nand_chip *chip = mtd_to_nand(mtd); struct mtd_ecc_stats stats; @@ -3855,6 +3856,8 @@ static int nand_do_read_oob(struct mtd_info *mtd, loff_t from, if (!readlen) break; + max_bitflips = max_t(unsigned int, max_bitflips, ret); + /* Increment page address */ realpage++; @@ -3876,7 +3879,7 @@ static int nand_do_read_oob(struct mtd_info *mtd, loff_t from, if (mtd->ecc_stats.failed - stats.failed) return -EBADMSG; - return mtd->ecc_stats.corrected - stats.corrected ? -EUCLEAN : 0; + return max_bitflips; } /** -- 2.11.0

7 years, 3 months

2
2
0 0

[Linux-stable-mirror] [PATCH v4] RFT: mmc: sdhci: Implement an SDHCI-specific bounce buffer

by Linus Walleij

The bounce buffer is gone from the MMC core, and now we found out that there are some (crippled) i.MX boards out there that have broken ADMA (cannot do scatter-gather), and broken PIO so they must use SDMA. Closer examination shows a less significant slowdown also on SDMA-only capable Laptop hosts. SDMA sets down the number of segments to one, so that each segment gets turned into a singular request that ping-pongs to the block layer before the next request/segment is issued. Apparently it happens a lot that the block layer send requests that include a lot of physically discontigous segments. My guess is that this phenomenon is coming from the file system. These devices that cannot handle scatterlists in hardware can see major benefits from a DMA-contigous bounce buffer. This patch accumulates those fragmented scatterlists in a physically contigous bounce buffer so that we can issue bigger DMA data chunks to/from the card. When tested with thise PCI-integrated host (1217:8221) that only supports SDMA: 0b:00.0 SD Host controller: O2 Micro, Inc. OZ600FJ0/OZ900FJ0/OZ600FJS SD/MMC Card Reader Controller (rev 05) This patch gave ~1Mbyte/s improved throughput on large reads and writes when testing using iozone than without the patch. On the i.MX SDHCI controllers on the crippled i.MX 25 and i.MX 35 the patch restores the performance to what it was before we removed the bounce buffers, and then some: performance is better than ever because we now allocate a bounce buffer the size of the maximum single request the SDMA engine can handle. On the PCI laptop this is 256K, whereas with the old bounce buffer code it was 64K max. Cc: Benjamin Beckmeyer <beckmeyer.b(a)rittal.de> Cc: Pierre Ossman <pierre(a)ossman.eu> Cc: Benoît Thébaudeau <benoit(a)wsystem.com> Cc: Fabio Estevam <fabio.estevam(a)nxp.com> Cc: stable(a)vger.kernel.org Fixes: de3ee99b097d ("mmc: Delete bounce buffer handling") Signed-off-by: Linus Walleij <linus.walleij(a)linaro.org> --- --- ChangeLog v3->v4: - Cap the bounce buffer to 64KB instead of the biggest segment as we experience diminishing returns with buffers > 64KB. - Instead of using dma_alloc_coherent(), use good old devm_kmalloc() and issue dma_sync_single_for*() to explicitly switch ownership between CPU and the device. This way we exercise the cache better and may consume less CPU. - Bail out with single segments if we cannot allocate a bounce buffer. - Tested on the PCI SDHCI on my laptop: requesting a new test on i.MX from Benjamin. (Please!) ChangeLog v2->v3: - Rewrite the commit message a bit - Add Benjamin's Tested-by - Add Fixes and stable tags ChangeLog v1->v2: - Skip the remapping and fiddling with the buffer, instead use dma_alloc_coherent() and use a simple, coherent bounce buffer. - Couple kernel messages to ->parent of the mmc_host as it relates to the hardware characteristics. --- drivers/mmc/host/sdhci.c | 125 ++++++++++++++++++++++++++++++++++++++++++++--- drivers/mmc/host/sdhci.h | 3 ++ 2 files changed, 120 insertions(+), 8 deletions(-) diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c index e9290a3439d5..694a320d9444 100644 --- a/drivers/mmc/host/sdhci.c +++ b/drivers/mmc/host/sdhci.c @@ -21,6 +21,7 @@ #include <linux/dma-mapping.h> #include <linux/slab.h> #include <linux/scatterlist.h> +#include <linux/sizes.h> #include <linux/swiotlb.h> #include <linux/regulator/consumer.h> #include <linux/pm_runtime.h> @@ -502,8 +503,27 @@ static int sdhci_pre_dma_transfer(struct sdhci_host *host, if (data->host_cookie == COOKIE_PRE_MAPPED) return data->sg_count; - sg_count = dma_map_sg(mmc_dev(host->mmc), data->sg, data->sg_len, - mmc_get_dma_dir(data)); + /* Bounce write requests to the bounce buffer */ + if (host->bounce_buffer) { + if (mmc_get_dma_dir(data) == DMA_TO_DEVICE) { + /* Copy the data to the bounce buffer */ + sg_copy_to_buffer(data->sg, data->sg_len, + host->bounce_buffer, + host->bounce_buffer_size); + } + /* Switch ownership to the DMA */ + dma_sync_single_for_device(host->mmc->parent, + host->bounce_addr, + host->bounce_buffer_size, + DMA_TO_DEVICE); + /* Just a dummy value */ + sg_count = 1; + } else { + /* Just access the data directly from memory */ + sg_count = dma_map_sg(mmc_dev(host->mmc), + data->sg, data->sg_len, + mmc_get_dma_dir(data)); + } if (sg_count == 0) return -ENOSPC; @@ -858,8 +878,13 @@ static void sdhci_prepare_data(struct sdhci_host *host, struct mmc_command *cmd) SDHCI_ADMA_ADDRESS_HI); } else { WARN_ON(sg_cnt != 1); - sdhci_writel(host, sg_dma_address(data->sg), - SDHCI_DMA_ADDRESS); + /* Bounce buffer goes to work */ + if (host->bounce_buffer) + sdhci_writel(host, host->bounce_addr, + SDHCI_DMA_ADDRESS); + else + sdhci_writel(host, sg_dma_address(data->sg), + SDHCI_DMA_ADDRESS); } } @@ -2248,7 +2273,12 @@ static void sdhci_pre_req(struct mmc_host *mmc, struct mmc_request *mrq) mrq->data->host_cookie = COOKIE_UNMAPPED; - if (host->flags & SDHCI_REQ_USE_DMA) + /* + * No pre-mapping in the pre hook if we're using the bounce buffer, + * for that we would need two bounce buffers since one buffer is + * in flight when this is getting called. + */ + if (host->flags & SDHCI_REQ_USE_DMA && !host->bounce_buffer) sdhci_pre_dma_transfer(host, mrq->data, COOKIE_PRE_MAPPED); } @@ -2352,8 +2382,28 @@ static bool sdhci_request_done(struct sdhci_host *host) struct mmc_data *data = mrq->data; if (data && data->host_cookie == COOKIE_MAPPED) { - dma_unmap_sg(mmc_dev(host->mmc), data->sg, data->sg_len, - mmc_get_dma_dir(data)); + if (host->bounce_buffer) { + /* + * On reads, copy the bounced data into the + * sglist + */ + if (mmc_get_dma_dir(data) == DMA_FROM_DEVICE) { + dma_sync_single_for_cpu( + host->mmc->parent, + host->bounce_addr, + host->bounce_buffer_size, + DMA_FROM_DEVICE); + sg_copy_from_buffer(data->sg, + data->sg_len, + host->bounce_buffer, + host->bounce_buffer_size); + } + } else { + /* Unmap the raw data */ + dma_unmap_sg(mmc_dev(host->mmc), data->sg, + data->sg_len, + mmc_get_dma_dir(data)); + } data->host_cookie = COOKIE_UNMAPPED; } } @@ -2636,7 +2686,12 @@ static void sdhci_data_irq(struct sdhci_host *host, u32 intmask) */ if (intmask & SDHCI_INT_DMA_END) { u32 dmastart, dmanow; - dmastart = sg_dma_address(host->data->sg); + + if (host->bounce_buffer) + dmastart = host->bounce_addr; + else + dmastart = sg_dma_address(host->data->sg); + dmanow = dmastart + host->data->bytes_xfered; /* * Force update to the next DMA block boundary. @@ -3713,6 +3768,60 @@ int sdhci_setup_host(struct sdhci_host *host) */ mmc->max_blk_count = (host->quirks & SDHCI_QUIRK_NO_MULTIBLOCK) ? 1 : 65535; + if (mmc->max_segs == 1) { + unsigned int max_blocks; + unsigned int max_seg_size; + + /* + * Cap the bounce buffer at 64KB. Using a bigger bounce buffer + * has diminishing returns, this is probably because SD/MMC + * cards are usually optimized to handle this size of requests. + */ + max_seg_size = SZ_64K; + if (mmc->max_req_size < max_seg_size) + max_seg_size = mmc->max_req_size; + max_blocks = max_seg_size / 512; + dev_info(mmc->parent, + "host only supports SDMA, activate bounce buffer\n"); + + /* + * When we just support one segment, we can get significant + * speedups by the help of a bounce buffer to group scattered + * reads/writes together. + */ + host->bounce_buffer = devm_kmalloc(mmc->parent, + max_seg_size, + GFP_KERNEL); + if (!host->bounce_buffer) { + dev_err(mmc->parent, + "failed to allocate %u bytes for bounce buffer, falling back to single segments\n", + max_seg_size); + /* + * Exiting with zero here makes sure we proceed with + * mmc->max_segs == 1. + */ + return 0; + } + + host->bounce_buffer_size = max_seg_size; + host->bounce_addr = dma_map_single(mmc->parent, + host->bounce_buffer, + host->bounce_buffer_size, + DMA_BIDIRECTIONAL); + ret = dma_mapping_error(mmc->parent, host->bounce_addr); + if (ret) + /* Again fall back to max_segs == 1 */ + return 0; + + /* Lie about this since we're bouncing */ + mmc->max_segs = max_blocks; + mmc->max_seg_size = max_seg_size; + + dev_info(mmc->parent, + "bounce buffer: bounce up to %u segments into one, max segment size %u bytes\n", + max_blocks, max_seg_size); + } + return 0; unreg: diff --git a/drivers/mmc/host/sdhci.h b/drivers/mmc/host/sdhci.h index 54bc444c317f..865e09618d22 100644 --- a/drivers/mmc/host/sdhci.h +++ b/drivers/mmc/host/sdhci.h @@ -440,6 +440,9 @@ struct sdhci_host { int irq; /* Device IRQ */ void __iomem *ioaddr; /* Mapped address */ + char *bounce_buffer; /* For packing SDMA reads/writes */ + dma_addr_t bounce_addr; + size_t bounce_buffer_size; const struct sdhci_ops *ops; /* Low level hw interface */ -- 2.14.3

7 years, 3 months

2
2
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror