From patchwork Wed Sep 18 12:59:08 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: =?utf-8?q?Thomas_Hellstr=C3=B6m_=28Intel=29?= X-Patchwork-Id: 11150319 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 33BDF912 for ; Wed, 18 Sep 2019 12:59:47 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 1C2D721924 for ; Wed, 18 Sep 2019 12:59:47 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 1C2D721924 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=shipmail.org Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=dri-devel-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 8664689C54; Wed, 18 Sep 2019 12:59:43 +0000 (UTC) X-Original-To: dri-devel@lists.freedesktop.org Delivered-To: dri-devel@lists.freedesktop.org Received: from pio-pvt-msa2.bahnhof.se (pio-pvt-msa2.bahnhof.se [79.136.2.41]) by gabe.freedesktop.org (Postfix) with ESMTPS id 447876EF34 for ; Wed, 18 Sep 2019 12:59:35 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by pio-pvt-msa2.bahnhof.se (Postfix) with ESMTP id 52BE43FBB0; Wed, 18 Sep 2019 14:59:28 +0200 (CEST) X-Virus-Scanned: Debian amavisd-new at bahnhof.se X-Spam-Flag: NO X-Spam-Score: -2.099 X-Spam-Level: X-Spam-Status: No, score=-2.099 tagged_above=-999 required=6.31 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no Received: from pio-pvt-msa2.bahnhof.se ([127.0.0.1]) by localhost (pio-pvt-msa2.bahnhof.se [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id J6g-nBY-BBV3; Wed, 18 Sep 2019 14:59:27 +0200 (CEST) Received: from mail1.shipmail.org (h-205-35.A357.priv.bahnhof.se [155.4.205.35]) (Authenticated sender: mb878879) by pio-pvt-msa2.bahnhof.se (Postfix) with ESMTPA id AAB473F746; Wed, 18 Sep 2019 14:59:24 +0200 (CEST) Received: from localhost.localdomain.localdomain (h-205-35.A357.priv.bahnhof.se [155.4.205.35]) by mail1.shipmail.org (Postfix) with ESMTPSA id 18D413602D9; Wed, 18 Sep 2019 14:59:24 +0200 (CEST) From: =?utf-8?q?Thomas_Hellstr=C3=B6m_=28VMware=29?= To: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, linux-mm@kvack.org Subject: [PATCH 1/7] mm: Add write-protect and clean utilities for address space ranges Date: Wed, 18 Sep 2019 14:59:08 +0200 Message-Id: <20190918125914.38497-2-thomas_os@shipmail.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20190918125914.38497-1-thomas_os@shipmail.org> References: <20190918125914.38497-1-thomas_os@shipmail.org> MIME-Version: 1.0 X-Mailman-Original-DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=shipmail.org; s=mail; t=1568811564; bh=8endBo4LYeRXjKpQW+eclcRDwX2DG11DoosCTAG8UE8=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=L9GUFdT6PefsBIiZsfCr9qPb1w/PpHYQ2WRSZi1zMMn55SxVWMIdnNaO7Vbs5y0qP qfN0To//3d1Rit4sPvLhpO15DQnew7ORKTsQrwVjlJALXDaqs3pJaO2aTIzZMjUzsY e+1uzuDjV2GxMtg49NJOxESG8d8iVnRQBq6Vccso= X-Mailman-Original-Authentication-Results: pio-pvt-msa2.bahnhof.se; dkim=pass (1024-bit key; unprotected) header.d=shipmail.org header.i=@shipmail.org header.b=L9GUFdT6; dkim-atps=neutral X-BeenThere: dri-devel@lists.freedesktop.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Direct Rendering Infrastructure - Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Thomas Hellstrom , Michal Hocko , Rik van Riel , pv-drivers@vmware.com, Minchan Kim , Will Deacon , Ralph Campbell , Matthew Wilcox , Peter Zijlstra , =?utf-8?b?SsOpcsO0bWUgR2xpc3Nl?= , linux-graphics-maintainer@vmware.com, Souptick Joarder , Huang Ying , Andrew Morton Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" From: Thomas Hellstrom Add two utilities to a) write-protect and b) clean all ptes pointing into a range of an address space. The utilities are intended to aid in tracking dirty pages (either driver-allocated system memory or pci device memory). The write-protect utility should be used in conjunction with page_mkwrite() and pfn_mkwrite() to trigger write page-faults on page accesses. Typically one would want to use this on sparse accesses into large memory regions. The clean utility should be used to utilize hardware dirtying functionality and avoid the overhead of page-faults, typically on large accesses into small memory regions. The added file "as_dirty_helpers.c" is initially listed as maintained by VMware under our DRM driver. If somebody would like it elsewhere, that's of course no problem. Cc: Andrew Morton Cc: Matthew Wilcox Cc: Will Deacon Cc: Peter Zijlstra Cc: Rik van Riel Cc: Minchan Kim Cc: Michal Hocko Cc: Huang Ying Cc: Souptick Joarder Cc: "Jérôme Glisse" Cc: linux-mm@kvack.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Thomas Hellstrom Reviewed-by: Ralph Campbell #v1 --- MAINTAINERS | 1 + include/linux/mm.h | 13 +- mm/Kconfig | 3 + mm/Makefile | 1 + mm/as_dirty_helpers.c | 392 ++++++++++++++++++++++++++++++++++++++++++ 5 files changed, 409 insertions(+), 1 deletion(-) create mode 100644 mm/as_dirty_helpers.c diff --git a/MAINTAINERS b/MAINTAINERS index c2d975da561f..b596c7cf4a85 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -5287,6 +5287,7 @@ T: git git://people.freedesktop.org/~thomash/linux S: Supported F: drivers/gpu/drm/vmwgfx/ F: include/uapi/drm/vmwgfx_drm.h +F: mm/as_dirty_helpers.c DRM DRIVERS M: David Airlie diff --git a/include/linux/mm.h b/include/linux/mm.h index 0334ca97c584..27ff341ecbdc 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -2657,7 +2657,6 @@ typedef int (*pte_fn_t)(pte_t *pte, unsigned long addr, void *data); extern int apply_to_page_range(struct mm_struct *mm, unsigned long address, unsigned long size, pte_fn_t fn, void *data); - #ifdef CONFIG_PAGE_POISONING extern bool page_poisoning_enabled(void); extern void kernel_poison_pages(struct page *page, int numpages, int enable); @@ -2891,5 +2890,17 @@ void __init setup_nr_node_ids(void); static inline void setup_nr_node_ids(void) {} #endif +#ifdef CONFIG_AS_DIRTY_HELPERS +unsigned long apply_as_clean(struct address_space *mapping, + pgoff_t first_index, pgoff_t nr, + pgoff_t bitmap_pgoff, + unsigned long *bitmap, + pgoff_t *start, + pgoff_t *end); + +unsigned long apply_as_wrprotect(struct address_space *mapping, + pgoff_t first_index, pgoff_t nr); +#endif + #endif /* __KERNEL__ */ #endif /* _LINUX_MM_H */ diff --git a/mm/Kconfig b/mm/Kconfig index 56cec636a1fc..594350e9d78e 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -736,4 +736,7 @@ config ARCH_HAS_PTE_SPECIAL config ARCH_HAS_HUGEPD bool +config AS_DIRTY_HELPERS + bool + endmenu diff --git a/mm/Makefile b/mm/Makefile index d0b295c3b764..4086f1eefbc6 100644 --- a/mm/Makefile +++ b/mm/Makefile @@ -105,3 +105,4 @@ obj-$(CONFIG_PERCPU_STATS) += percpu-stats.o obj-$(CONFIG_ZONE_DEVICE) += memremap.o obj-$(CONFIG_HMM_MIRROR) += hmm.o obj-$(CONFIG_MEMFD_CREATE) += memfd.o +obj-$(CONFIG_AS_DIRTY_HELPERS) += as_dirty_helpers.o diff --git a/mm/as_dirty_helpers.c b/mm/as_dirty_helpers.c new file mode 100644 index 000000000000..d4cc37dcb144 --- /dev/null +++ b/mm/as_dirty_helpers.c @@ -0,0 +1,392 @@ +// SPDX-License-Identifier: GPL-2.0 +#include +#include +#include +#include +#include +#include +#include + +/** + * struct as_walk - Argument to struct as_walk_ops callbacks. + * @vma: Pointer to the struct vmw_area_struct currently being walked. + * + * Embeddable argument to struct as_walk_ops callbacks. + */ +struct as_walk { + struct vm_area_struct *vma; +}; + +/** + * struct as_walk_ops - Callbacks for entries of various page table levels. + * extend for additional level support. + */ +struct as_walk_ops { + /** + * pte-entry: Callback for PTEs + * @pte: Pointer to the PTE. + * @addr: Virtual address. + * @asw: Struct as_walk argument for the walk. Embed for additional + * data. + */ + void (*const pte_entry) (pte_t *pte, unsigned long addr, + struct as_walk *asw); +}; + +/* Page-walking code */ +static void walk_as_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end, + const struct as_walk_ops *ops, + struct as_walk *asw) +{ + struct mm_struct *mm = asw->vma->vm_mm; + pte_t *pte; + spinlock_t *ptl; + + pte = (mm == &init_mm) ? + pte_offset_kernel(pmd, addr) : + pte_offset_map_lock(mm, pmd, addr, &ptl); + + arch_enter_lazy_mmu_mode(); + + do { + ops->pte_entry(pte++, addr, asw); + } while (addr += PAGE_SIZE, addr != end); + + arch_leave_lazy_mmu_mode(); + + if (mm != &init_mm) + pte_unmap_unlock(pte - 1, ptl); +} + +static void walk_as_pmd_range(pud_t *pud, unsigned long addr, unsigned long end, + const struct as_walk_ops *ops, + struct as_walk *asw) +{ + pmd_t *pmd = pmd_offset(pud, addr); + unsigned long next; + + do { + next = pmd_addr_end(addr, end); + if (pmd_none_or_clear_bad(pmd)) + continue; + if (WARN_ON(pmd_huge(*pmd))) + continue; + walk_as_pte_range(pmd, addr, next, ops, asw); + } while (pmd++, addr = next, addr != end); +} + +static void walk_as_pud_range(p4d_t *p4d, unsigned long addr, unsigned long end, + const struct as_walk_ops *ops, + struct as_walk *asw) +{ + pud_t *pud = pud_offset(p4d, addr); + unsigned long next; + + do { + next = pud_addr_end(addr, end); + if (pud_none_or_clear_bad(pud)) + continue; + if (WARN_ON(pud_huge(*pud))) + continue; + walk_as_pmd_range(pud, addr, next, ops, asw); + } while (pud++, addr = next, addr != end); +} + +static void walk_as_p4d_range(pgd_t *pgd, unsigned long addr, unsigned long end, + const struct as_walk_ops *ops, + struct as_walk *asw) +{ + p4d_t *p4d = p4d_offset(pgd, addr); + unsigned long next; + + do { + next = p4d_addr_end(addr, end); + if (p4d_none_or_clear_bad(p4d)) + continue; + walk_as_pud_range(p4d, addr, next, ops, asw); + } while (p4d++, addr = next, addr != end); +} + +static void walk_as_pfn_range(unsigned long addr, unsigned long end, + const struct as_walk_ops *ops, + struct as_walk *asw) +{ + pgd_t *pgd = pgd_offset(asw->vma->vm_mm, addr); + unsigned long next; + + do { + next = pgd_addr_end(addr, end); + if (pgd_none_or_clear_bad(pgd)) + continue; + walk_as_p4d_range(pgd, addr, next, ops, asw); + } while (pgd++, addr = next, addr != end); +} + + +/** + * struct as_walk_range - Argument for apply_as_range + * @asw: The struct as_walk we embed for the page walk + * @start: Address of first modified pte + * @end: Address of last modified pte + 1 + * @total: Total number of modified ptes + */ +struct as_walk_range { + struct as_walk base; + unsigned long start; + unsigned long end; + unsigned long total; +}; + +#define to_as_walk_range(_asw) container_of(_asw, struct as_walk_range, base) + +/** + * apply_pt_wrprotect - Leaf pte callback to write-protect a pte + * @pte: Pointer to the pte + * @addr: The virtual page address + * @asw: Pointer to a struct as_walk embedded in a struct as_walk_range + * + * The function write-protects a pte and records the range in + * virtual address space of touched ptes for efficient range TLB flushes. + */ +static void apply_pt_wrprotect(pte_t *pte, unsigned long addr, + struct as_walk *asw) +{ + struct as_walk_range *awr = to_as_walk_range(asw); + pte_t ptent = *pte; + + if (pte_write(ptent)) { + pte_t old_pte = ptep_modify_prot_start(asw->vma, addr, pte); + + ptent = pte_wrprotect(old_pte); + ptep_modify_prot_commit(asw->vma, addr, pte, old_pte, ptent); + awr->total++; + awr->start = min(awr->start, addr); + awr->end = max(awr->end, addr + PAGE_SIZE); + } +} + +/** + * struct as_walk_clean - Argument structure for apply_pt_clean + * @base: struct as_walk we derive from + * @bitmap_pgoff: Address_space Page offset of the first bit in @bitmap + * @bitmap: Bitmap with one bit for each page offset in the address_space range + * covered. + * @start: Address_space page offset of first modified pte relative + * to @bitmap_pgoff + * @end: Address_space page offset of last modified pte relative + * to @bitmap_pgoff + */ +struct as_walk_clean { + struct as_walk_range base; + pgoff_t bitmap_pgoff; + unsigned long *bitmap; + pgoff_t start; + pgoff_t end; +}; + +#define to_as_walk_clean(_awr) container_of(_awr, struct as_walk_clean, base) + +/** + * apply_pt_clean - Leaf pte callback to clean a pte + * @pte: Pointer to the pte + * @addr: The virtual page address + * @asw: Pointer to a struct as_walk embedded in a struct as_walk_clean + * + * The function cleans a pte and records the range in + * virtual address space of touched ptes for efficient TLB flushes. + * It also records dirty ptes in a bitmap representing page offsets + * in the address_space, as well as the first and last of the bits + * touched. + */ +static void apply_pt_clean(pte_t *pte, unsigned long addr, struct as_walk *asw) +{ + struct as_walk_range *awr = to_as_walk_range(asw); + struct as_walk_clean *clean = to_as_walk_clean(awr); + pte_t ptent = *pte; + + if (pte_dirty(ptent)) { + pgoff_t pgoff = ((addr - asw->vma->vm_start) >> PAGE_SHIFT) + + asw->vma->vm_pgoff - clean->bitmap_pgoff; + pte_t old_pte = ptep_modify_prot_start(asw->vma, addr, pte); + + ptent = pte_mkclean(old_pte); + ptep_modify_prot_commit(asw->vma, addr, pte, old_pte, ptent); + + awr->total++; + awr->start = min(awr->start, addr); + awr->end = max(awr->end, addr + PAGE_SIZE); + + __set_bit(pgoff, clean->bitmap); + clean->start = min(clean->start, pgoff); + clean->end = max(clean->end, pgoff + 1); + } +} + +/** + * apply_as_range - Apply a pte callback to all PTEs pointing into a range + * of an address_space. + * @mapping: Pointer to the struct address_space + * @aas: Closure structure + * @first_index: First page offset in the address_space + * @nr: Number of incremental page offsets to cover + * + * Return: Number of ptes touched. Note that this number might be larger + * than @nr if there are overlapping vmas + */ +static unsigned long apply_as_range(struct address_space *mapping, + pgoff_t first_index, pgoff_t nr, + const struct as_walk_ops *ops, + struct as_walk_range *awr) +{ + struct vm_area_struct *vma; + pgoff_t vba, vea, cba, cea; + unsigned long start_addr, end_addr; + struct mmu_notifier_range range; + + i_mmap_lock_read(mapping); + vma_interval_tree_foreach(vma, &mapping->i_mmap, first_index, + first_index + nr - 1) { + unsigned long vm_flags = READ_ONCE(vma->vm_flags); + + /* + * We can only do advisory flag tests below, since we can't + * require the mm's mmap_sem to be held to protect the flags. + * Therefore, callers that strictly depend on specific vm_flags + * to remain constant throughout the operation must ensure + * those flags are immutable for all relevant vmas or can't use + * this function. Fixing this properly would require the + * vm_flags to be protected by a separate lock taken after the + * i_mmap_lock + */ + + /* Skip non-applicable VMAs */ + if ((vm_flags & (VM_SHARED | VM_WRITE)) != + (VM_SHARED | VM_WRITE)) + continue; + + /* Warn on and skip VMAs whose flags indicate illegal usage */ + if (WARN_ON((vm_flags & (VM_HUGETLB | VM_IO)) != VM_IO)) + continue; + + /* Clip to the vma */ + vba = vma->vm_pgoff; + vea = vba + vma_pages(vma); + cba = first_index; + cba = max(cba, vba); + cea = first_index + nr; + cea = min(cea, vea); + + /* Translate to virtual address */ + start_addr = ((cba - vba) << PAGE_SHIFT) + vma->vm_start; + end_addr = ((cea - vba) << PAGE_SHIFT) + vma->vm_start; + if (start_addr >= end_addr) + continue; + + awr->base.vma = vma; + awr->start = end_addr; + awr->end = start_addr; + + mmu_notifier_range_init(&range, MMU_NOTIFY_PROTECTION_PAGE, 0, + vma, vma->vm_mm, start_addr, end_addr); + mmu_notifier_invalidate_range_start(&range); + + /* Is this needed when we only change protection? */ + flush_cache_range(vma, start_addr, end_addr); + + /* + * We're not using tlb_gather_mmu() since typically + * only a small subrange of PTEs are affected, whereas + * tlb_gather_mmu() records the full range. + */ + inc_tlb_flush_pending(vma->vm_mm); + walk_as_pfn_range(start_addr, end_addr, ops, &awr->base); + if (awr->end > awr->start) + flush_tlb_range(vma, awr->start, awr->end); + + mmu_notifier_invalidate_range_end(&range); + dec_tlb_flush_pending(vma->vm_mm); + } + i_mmap_unlock_read(mapping); + + return awr->total; +} + +/** + * apply_as_wrprotect - Write-protect all ptes in an address_space range + * @mapping: The address_space we want to write protect + * @first_index: The first page offset in the range + * @nr: Number of incremental page offsets to cover + * + * WARNING: This function should only be used for address spaces whose + * vmas are marked VM_IO and that do not contain huge pages. + * To avoid interference with COW'd pages, vmas not marked VM_SHARED are + * simply skipped. + * + * Return: The number of ptes actually write-protected. Note that + * already write-protected ptes are not counted. + */ +unsigned long apply_as_wrprotect(struct address_space *mapping, + pgoff_t first_index, pgoff_t nr) +{ + static const struct as_walk_ops ops = { + .pte_entry = apply_pt_wrprotect + }; + struct as_walk_range awr = { .total = 0 }; + + return apply_as_range(mapping, first_index, nr, &ops, &awr); +} +EXPORT_SYMBOL_GPL(apply_as_wrprotect); + +/** + * apply_as_clean - Clean all ptes in an address_space range + * @mapping: The address_space we want to clean + * @first_index: The first page offset in the range + * @nr: Number of incremental page offsets to cover + * @bitmap_pgoff: The page offset of the first bit in @bitmap + * @bitmap: Pointer to a bitmap of at least @nr bits. The bitmap needs to + * cover the whole range @first_index..@first_index + @nr. + * @start: Pointer to number of the first set bit in @bitmap. + * is modified as new bits are set by the function. + * @end: Pointer to the number of the last set bit in @bitmap. + * none set. The value is modified as new bits are set by the function. + * + * Note: When this function returns there is no guarantee that a CPU has + * not already dirtied new ptes. However it will not clean any ptes not + * reported in the bitmap. + * + * If a caller needs to make sure all dirty ptes are picked up and none + * additional are added, it first needs to write-protect the address-space + * range and make sure new writers are blocked in page_mkwrite() or + * pfn_mkwrite(). And then after a TLB flush following the write-protection + * pick up all dirty bits. + * + * WARNING: This function should only be used for address spaces whose + * vmas are marked VM_IO and that do not contain huge pages. + * To avoid interference with COW'd pages, vmas not marked VM_SHARED are + * simply skipped. + * + * Return: The number of dirty ptes actually cleaned. + */ +unsigned long apply_as_clean(struct address_space *mapping, + pgoff_t first_index, pgoff_t nr, + pgoff_t bitmap_pgoff, + unsigned long *bitmap, + pgoff_t *start, + pgoff_t *end) +{ + bool none_set = (*start >= *end); + static const struct as_walk_ops ops = { .pte_entry = apply_pt_clean }; + struct as_walk_clean clean = { + .base = { .total = 0, }, + .bitmap_pgoff = bitmap_pgoff, + .bitmap = bitmap, + .start = none_set ? nr : *start, + .end = none_set ? 0 : *end, + }; + unsigned long ret = apply_as_range(mapping, first_index, nr, &ops, + &clean.base); + *start = clean.start; + *end = clean.end; + return ret; +} +EXPORT_SYMBOL_GPL(apply_as_clean); From patchwork Wed Sep 18 12:59:09 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: =?utf-8?q?Thomas_Hellstr=C3=B6m_=28Intel=29?= X-Patchwork-Id: 11150301 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 0734B912 for ; Wed, 18 Sep 2019 12:59:34 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id E3A8321924 for ; Wed, 18 Sep 2019 12:59:33 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org E3A8321924 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=shipmail.org Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=dri-devel-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 59EE96EF32; Wed, 18 Sep 2019 12:59:31 +0000 (UTC) X-Original-To: dri-devel@lists.freedesktop.org Delivered-To: dri-devel@lists.freedesktop.org Received: from pio-pvt-msa1.bahnhof.se (pio-pvt-msa1.bahnhof.se [79.136.2.40]) by gabe.freedesktop.org (Postfix) with ESMTPS id 9F59F6EF31 for ; Wed, 18 Sep 2019 12:59:30 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by pio-pvt-msa1.bahnhof.se (Postfix) with ESMTP id CA0663F51E; Wed, 18 Sep 2019 14:59:28 +0200 (CEST) X-Virus-Scanned: Debian amavisd-new at bahnhof.se X-Spam-Flag: NO X-Spam-Score: -2.099 X-Spam-Level: X-Spam-Status: No, score=-2.099 tagged_above=-999 required=6.31 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no Received: from pio-pvt-msa1.bahnhof.se ([127.0.0.1]) by localhost (pio-pvt-msa1.bahnhof.se [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ie7I6TOX5u-9; Wed, 18 Sep 2019 14:59:28 +0200 (CEST) Received: from mail1.shipmail.org (h-205-35.A357.priv.bahnhof.se [155.4.205.35]) (Authenticated sender: mb878879) by pio-pvt-msa1.bahnhof.se (Postfix) with ESMTPA id BB6063F549; Wed, 18 Sep 2019 14:59:24 +0200 (CEST) Received: from localhost.localdomain.localdomain (h-205-35.A357.priv.bahnhof.se [155.4.205.35]) by mail1.shipmail.org (Postfix) with ESMTPSA id 4918A36031D; Wed, 18 Sep 2019 14:59:24 +0200 (CEST) From: =?utf-8?q?Thomas_Hellstr=C3=B6m_=28VMware=29?= To: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, linux-mm@kvack.org Subject: [PATCH 2/7] drm/ttm: Remove explicit typecasts of vm_private_data Date: Wed, 18 Sep 2019 14:59:09 +0200 Message-Id: <20190918125914.38497-3-thomas_os@shipmail.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20190918125914.38497-1-thomas_os@shipmail.org> References: <20190918125914.38497-1-thomas_os@shipmail.org> MIME-Version: 1.0 X-Mailman-Original-DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=shipmail.org; s=mail; t=1568811564; bh=R7so/IGdQQ1rwaSJyPiZqti40YJDt90potOk+4jpddE=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=r9L3bfmr6V+Rgk+/d++uKtKktRO+85SKrG+zpdhKK0NyALMkfvZMsfWA/Um/9pYwK 7VFgVgxC2m/jU3T7YWAAfbhU5mkr0pA5iNJI/yTmKf+nxfpjl7ZzcPR/Lb9LevLQTF 9zn1OzTI2iY4ewmuPOSaceriTaNSK4UajXVjU6Uw= X-Mailman-Original-Authentication-Results: pio-pvt-msa1.bahnhof.se; dkim=pass (1024-bit key; unprotected) header.d=shipmail.org header.i=@shipmail.org header.b="r9L3bfmr"; dkim-atps=neutral X-BeenThere: dri-devel@lists.freedesktop.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Direct Rendering Infrastructure - Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: pv-drivers@vmware.com, Thomas Hellstrom , linux-graphics-maintainer@vmware.com, Matthew Wilcox Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" From: Thomas Hellstrom The explicit typcasts are meaningless, so remove them. Suggested-by: Matthew Wilcox Signed-off-by: Thomas Hellstrom --- drivers/gpu/drm/ttm/ttm_bo_vm.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c index 76eedb963693..8963546bf245 100644 --- a/drivers/gpu/drm/ttm/ttm_bo_vm.c +++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c @@ -109,8 +109,7 @@ static unsigned long ttm_bo_io_mem_pfn(struct ttm_buffer_object *bo, static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf) { struct vm_area_struct *vma = vmf->vma; - struct ttm_buffer_object *bo = (struct ttm_buffer_object *) - vma->vm_private_data; + struct ttm_buffer_object *bo = vma->vm_private_data; struct ttm_bo_device *bdev = bo->bdev; unsigned long page_offset; unsigned long page_last; @@ -302,8 +301,7 @@ static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf) static void ttm_bo_vm_open(struct vm_area_struct *vma) { - struct ttm_buffer_object *bo = - (struct ttm_buffer_object *)vma->vm_private_data; + struct ttm_buffer_object *bo = vma->vm_private_data; WARN_ON(bo->bdev->dev_mapping != vma->vm_file->f_mapping); @@ -312,7 +310,7 @@ static void ttm_bo_vm_open(struct vm_area_struct *vma) static void ttm_bo_vm_close(struct vm_area_struct *vma) { - struct ttm_buffer_object *bo = (struct ttm_buffer_object *)vma->vm_private_data; + struct ttm_buffer_object *bo = vma->vm_private_data; ttm_bo_put(bo); vma->vm_private_data = NULL; From patchwork Wed Sep 18 12:59:10 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: =?utf-8?q?Thomas_Hellstr=C3=B6m_=28Intel=29?= X-Patchwork-Id: 11150305 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 407A1912 for ; Wed, 18 Sep 2019 12:59:37 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 286F721924 for ; Wed, 18 Sep 2019 12:59:36 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 286F721924 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=shipmail.org Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=dri-devel-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 569D86EF33; Wed, 18 Sep 2019 12:59:34 +0000 (UTC) X-Original-To: dri-devel@lists.freedesktop.org Delivered-To: dri-devel@lists.freedesktop.org Received: from ste-pvt-msa1.bahnhof.se (ste-pvt-msa1.bahnhof.se [213.80.101.70]) by gabe.freedesktop.org (Postfix) with ESMTPS id ABB1A6EF31 for ; Wed, 18 Sep 2019 12:59:31 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by ste-pvt-msa1.bahnhof.se (Postfix) with ESMTP id 30CD23F6DC; Wed, 18 Sep 2019 14:59:30 +0200 (CEST) X-Virus-Scanned: Debian amavisd-new at bahnhof.se X-Spam-Flag: NO X-Spam-Score: -2.099 X-Spam-Level: X-Spam-Status: No, score=-2.099 tagged_above=-999 required=6.31 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no Received: from ste-pvt-msa1.bahnhof.se ([127.0.0.1]) by localhost (ste-pvt-msa1.bahnhof.se [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id xiRUd--qyrav; Wed, 18 Sep 2019 14:59:28 +0200 (CEST) Received: from mail1.shipmail.org (h-205-35.A357.priv.bahnhof.se [155.4.205.35]) (Authenticated sender: mb878879) by ste-pvt-msa1.bahnhof.se (Postfix) with ESMTPA id 4F6C33F247; Wed, 18 Sep 2019 14:59:24 +0200 (CEST) Received: from localhost.localdomain.localdomain (h-205-35.A357.priv.bahnhof.se [155.4.205.35]) by mail1.shipmail.org (Postfix) with ESMTPSA id 6407E360326; Wed, 18 Sep 2019 14:59:24 +0200 (CEST) From: =?utf-8?q?Thomas_Hellstr=C3=B6m_=28VMware=29?= To: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, linux-mm@kvack.org Subject: [PATCH 3/7] drm/ttm: Convert vm callbacks to helpers Date: Wed, 18 Sep 2019 14:59:10 +0200 Message-Id: <20190918125914.38497-4-thomas_os@shipmail.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20190918125914.38497-1-thomas_os@shipmail.org> References: <20190918125914.38497-1-thomas_os@shipmail.org> MIME-Version: 1.0 X-Mailman-Original-DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=shipmail.org; s=mail; t=1568811564; bh=uMSGH2llUf0g/5IIoIoj2eeW+M0du8bxg9TfkDKQh5Q=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=Aru0W2Ygo/WXRRZlOXyYkAMlfPAOVVMc5lsIlkOv3UNx/D+Dek3O8RNELJxDKOa7q Ox4U2io+QoTysr3p3CYhqcsHuXCJK55HFiA7wCySzYpLvxFwUMGnayrPwYv+whlR4y +QDD3Cx+btbMqTy5FIa7yooKowVp9D9+rbshP3Z8= X-Mailman-Original-Authentication-Results: ste-pvt-msa1.bahnhof.se; dkim=pass (1024-bit key; unprotected) header.d=shipmail.org header.i=@shipmail.org header.b=Aru0W2Yg; dkim-atps=neutral X-BeenThere: dri-devel@lists.freedesktop.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Direct Rendering Infrastructure - Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Thomas Hellstrom , Michal Hocko , Rik van Riel , pv-drivers@vmware.com, Minchan Kim , Will Deacon , Matthew Wilcox , Christoph Hellwig , Peter Zijlstra , =?utf-8?b?SsOpcsO0bWUgR2xpc3Nl?= , linux-graphics-maintainer@vmware.com, Souptick Joarder , Huang Ying , Andrew Morton , =?utf-8?q?Christian_K=C3=B6nig?= Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" From: Thomas Hellstrom With the vmwgfx dirty tracking, the default TTM fault handler is not completely sufficient (vmwgfx need to modify the vma->vm_flags member, and also needs to restrict the number of prefaults). We also want to replicate the new ttm_bo_vm_reserve() functionality So start turning the TTM vm code into helpers: ttm_bo_vm_fault_reserved() and ttm_bo_vm_reserve(), and provide a default TTM fault handler for other drivers to use. Cc: Andrew Morton Cc: Matthew Wilcox Cc: Will Deacon Cc: Peter Zijlstra Cc: Rik van Riel Cc: Minchan Kim Cc: Michal Hocko Cc: Huang Ying Cc: Souptick Joarder Cc: "Jérôme Glisse" Cc: "Christian König" Cc: Christoph Hellwig Signed-off-by: Thomas Hellstrom --- drivers/gpu/drm/ttm/ttm_bo_vm.c | 168 ++++++++++++++++++++------------ include/drm/ttm/ttm_bo_api.h | 14 +++ 2 files changed, 119 insertions(+), 63 deletions(-) diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c index 8963546bf245..0ac11837533f 100644 --- a/drivers/gpu/drm/ttm/ttm_bo_vm.c +++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c @@ -42,8 +42,6 @@ #include #include -#define TTM_BO_VM_NUM_PREFAULT 16 - static vm_fault_t ttm_bo_vm_fault_idle(struct ttm_buffer_object *bo, struct vm_fault *vmf) { @@ -106,24 +104,30 @@ static unsigned long ttm_bo_io_mem_pfn(struct ttm_buffer_object *bo, + page_offset; } -static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf) +/** + * ttm_bo_vm_reserve - Reserve a buffer object in a retryable vm callback + * @bo: The buffer object + * @vmf: The fault structure handed to the callback + * + * vm callbacks like fault() and *_mkwrite() allow for the mm_sem to be dropped + * during long waits, and after the wait the callback will be restarted. This + * is to allow other threads using the same virtual memory space concurrent + * access to map(), unmap() completely unrelated buffer objects. TTM buffer + * object reservations sometimes wait for GPU and should therefore be + * considered long waits. This function reserves the buffer object interruptibly + * taking this into account. Starvation is avoided by the vm system not + * allowing too many repeated restarts. + * This function is intended to be used in customized fault() and _mkwrite() + * handlers. + * + * Return: + * 0 on success and the bo was reserved. + * VM_FAULT_RETRY if blocking wait. + * VM_FAULT_NOPAGE if blocking wait and retrying was not allowed. + */ +vm_fault_t ttm_bo_vm_reserve(struct ttm_buffer_object *bo, + struct vm_fault *vmf) { - struct vm_area_struct *vma = vmf->vma; - struct ttm_buffer_object *bo = vma->vm_private_data; - struct ttm_bo_device *bdev = bo->bdev; - unsigned long page_offset; - unsigned long page_last; - unsigned long pfn; - struct ttm_tt *ttm = NULL; - struct page *page; - int err; - int i; - vm_fault_t ret = VM_FAULT_NOPAGE; - unsigned long address = vmf->address; - struct ttm_mem_type_manager *man = - &bdev->man[bo->mem.mem_type]; - struct vm_area_struct cvma; - /* * Work around locking order reversal in fault / nopfn * between mmap_sem and bo_reserve: Perform a trylock operation @@ -150,14 +154,54 @@ static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf) return VM_FAULT_NOPAGE; } + return 0; +} +EXPORT_SYMBOL(ttm_bo_vm_reserve); + +/** + * ttm_bo_vm_fault_reserved - TTM fault helper + * @vmf: The struct vm_fault given as argument to the fault callback + * @prot: The page protection to be used for this memory area. + * @num_prefault: Maximum number of prefault pages. The caller may want to + * specify this based on madvice settings and the size of the GPU object + * backed by the memory. + * + * This function inserts one or more page table entries pointing to the + * memory backing the buffer object, and then returns a return code + * instructing the caller to retry the page access. + * + * Return: + * VM_FAULT_NOPAGE on success or pending signal + * VM_FAULT_SIGBUS on unspecified error + * VM_FAULT_OOM on out-of-memory + * VM_FAULT_RETRY if retryable wait + */ +vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf, + pgprot_t prot, + pgoff_t num_prefault) +{ + struct vm_area_struct *vma = vmf->vma; + struct vm_area_struct cvma = *vma; + struct ttm_buffer_object *bo = vma->vm_private_data; + struct ttm_bo_device *bdev = bo->bdev; + unsigned long page_offset; + unsigned long page_last; + unsigned long pfn; + struct ttm_tt *ttm = NULL; + struct page *page; + int err; + pgoff_t i; + vm_fault_t ret = VM_FAULT_NOPAGE; + unsigned long address = vmf->address; + struct ttm_mem_type_manager *man = + &bdev->man[bo->mem.mem_type]; + /* * Refuse to fault imported pages. This should be handled * (if at all) by redirecting mmap to the exporter. */ - if (bo->ttm && (bo->ttm->page_flags & TTM_PAGE_FLAG_SG)) { - ret = VM_FAULT_SIGBUS; - goto out_unlock; - } + if (bo->ttm && (bo->ttm->page_flags & TTM_PAGE_FLAG_SG)) + return VM_FAULT_SIGBUS; if (bdev->driver->fault_reserve_notify) { struct dma_fence *moving = dma_fence_get(bo->moving); @@ -168,11 +212,9 @@ static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf) break; case -EBUSY: case -ERESTARTSYS: - ret = VM_FAULT_NOPAGE; - goto out_unlock; + return VM_FAULT_NOPAGE; default: - ret = VM_FAULT_SIGBUS; - goto out_unlock; + return VM_FAULT_SIGBUS; } if (bo->moving != moving) { @@ -188,21 +230,12 @@ static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf) * move. */ ret = ttm_bo_vm_fault_idle(bo, vmf); - if (unlikely(ret != 0)) { - if (ret == VM_FAULT_RETRY && - !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) { - /* The BO has already been unreserved. */ - return ret; - } - - goto out_unlock; - } + if (unlikely(ret != 0)) + return ret; err = ttm_mem_io_lock(man, true); - if (unlikely(err != 0)) { - ret = VM_FAULT_NOPAGE; - goto out_unlock; - } + if (unlikely(err != 0)) + return VM_FAULT_NOPAGE; err = ttm_mem_io_reserve_vm(bo); if (unlikely(err != 0)) { ret = VM_FAULT_SIGBUS; @@ -219,18 +252,8 @@ static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf) goto out_io_unlock; } - /* - * Make a local vma copy to modify the page_prot member - * and vm_flags if necessary. The vma parameter is protected - * by mmap_sem in write mode. - */ - cvma = *vma; - cvma.vm_page_prot = vm_get_page_prot(cvma.vm_flags); - - if (bo->mem.bus.is_iomem) { - cvma.vm_page_prot = ttm_io_prot(bo->mem.placement, - cvma.vm_page_prot); - } else { + cvma.vm_page_prot = ttm_io_prot(bo->mem.placement, prot); + if (!bo->mem.bus.is_iomem) { struct ttm_operation_ctx ctx = { .interruptible = false, .no_wait_gpu = false, @@ -239,24 +262,21 @@ static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf) }; ttm = bo->ttm; - cvma.vm_page_prot = ttm_io_prot(bo->mem.placement, - cvma.vm_page_prot); - - /* Allocate all page at once, most common usage */ - if (ttm_tt_populate(ttm, &ctx)) { + if (ttm_tt_populate(bo->ttm, &ctx)) { ret = VM_FAULT_OOM; goto out_io_unlock; } + } else { + /* Iomem should not be marked encrypted */ + cvma.vm_page_prot = pgprot_decrypted(cvma.vm_page_prot); } /* * Speculatively prefault a number of pages. Only error on * first page. */ - for (i = 0; i < TTM_BO_VM_NUM_PREFAULT; ++i) { + for (i = 0; i < num_prefault; ++i) { if (bo->mem.bus.is_iomem) { - /* Iomem should not be marked encrypted */ - cvma.vm_page_prot = pgprot_decrypted(cvma.vm_page_prot); pfn = ttm_bo_io_mem_pfn(bo, page_offset); } else { page = ttm->pages[page_offset]; @@ -294,12 +314,32 @@ static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf) ret = VM_FAULT_NOPAGE; out_io_unlock: ttm_mem_io_unlock(man); -out_unlock: + return ret; +} +EXPORT_SYMBOL(ttm_bo_vm_fault_reserved); + +static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf) +{ + struct vm_area_struct *vma = vmf->vma; + pgprot_t prot; + struct ttm_buffer_object *bo = vma->vm_private_data; + vm_fault_t ret; + + ret = ttm_bo_vm_reserve(bo, vmf); + if (ret) + return ret; + + prot = vm_get_page_prot(vma->vm_flags); + ret = ttm_bo_vm_fault_reserved(vmf, prot, TTM_BO_VM_NUM_PREFAULT); + if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) + return ret; + dma_resv_unlock(bo->base.resv); + return ret; } -static void ttm_bo_vm_open(struct vm_area_struct *vma) +void ttm_bo_vm_open(struct vm_area_struct *vma) { struct ttm_buffer_object *bo = vma->vm_private_data; @@ -307,14 +347,16 @@ static void ttm_bo_vm_open(struct vm_area_struct *vma) ttm_bo_get(bo); } +EXPORT_SYMBOL(ttm_bo_vm_open); -static void ttm_bo_vm_close(struct vm_area_struct *vma) +void ttm_bo_vm_close(struct vm_area_struct *vma) { struct ttm_buffer_object *bo = vma->vm_private_data; ttm_bo_put(bo); vma->vm_private_data = NULL; } +EXPORT_SYMBOL(ttm_bo_vm_close); static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo, unsigned long offset, diff --git a/include/drm/ttm/ttm_bo_api.h b/include/drm/ttm/ttm_bo_api.h index 43c4929a2171..851260cbbb2f 100644 --- a/include/drm/ttm/ttm_bo_api.h +++ b/include/drm/ttm/ttm_bo_api.h @@ -785,4 +785,18 @@ static inline bool ttm_bo_uses_embedded_gem_object(struct ttm_buffer_object *bo) { return bo->base.dev != NULL; } + +/* Default number of pre-faulted pages in the TTM fault handler */ +#define TTM_BO_VM_NUM_PREFAULT 16 + +vm_fault_t ttm_bo_vm_reserve(struct ttm_buffer_object *bo, + struct vm_fault *vmf); + +vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf, + pgprot_t prot, + pgoff_t num_prefault); + +void ttm_bo_vm_open(struct vm_area_struct *vma); + +void ttm_bo_vm_close(struct vm_area_struct *vma); #endif From patchwork Wed Sep 18 12:59:11 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: =?utf-8?q?Thomas_Hellstr=C3=B6m_=28Intel=29?= X-Patchwork-Id: 11150327 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 5CF56912 for ; Wed, 18 Sep 2019 12:59:52 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 4536D207FC for ; Wed, 18 Sep 2019 12:59:52 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 4536D207FC Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=shipmail.org Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=dri-devel-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 283126EF3B; Wed, 18 Sep 2019 12:59:50 +0000 (UTC) X-Original-To: dri-devel@lists.freedesktop.org Delivered-To: dri-devel@lists.freedesktop.org Received: from ste-pvt-msa2.bahnhof.se (ste-pvt-msa2.bahnhof.se [213.80.101.71]) by gabe.freedesktop.org (Postfix) with ESMTPS id EF0C86EF3B for ; Wed, 18 Sep 2019 12:59:47 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by ste-pvt-msa2.bahnhof.se (Postfix) with ESMTP id 5E7EC3F81C; Wed, 18 Sep 2019 14:59:31 +0200 (CEST) X-Virus-Scanned: Debian amavisd-new at bahnhof.se X-Spam-Flag: NO X-Spam-Score: -2.099 X-Spam-Level: X-Spam-Status: No, score=-2.099 tagged_above=-999 required=6.31 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no Received: from ste-pvt-msa2.bahnhof.se ([127.0.0.1]) by localhost (ste-ftg-msa2.bahnhof.se [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id IjzORh8W1YJM; Wed, 18 Sep 2019 14:59:25 +0200 (CEST) Received: from mail1.shipmail.org (h-205-35.A357.priv.bahnhof.se [155.4.205.35]) (Authenticated sender: mb878879) by ste-pvt-msa2.bahnhof.se (Postfix) with ESMTPA id E91E83F610; Wed, 18 Sep 2019 14:59:24 +0200 (CEST) Received: from localhost.localdomain.localdomain (h-205-35.A357.priv.bahnhof.se [155.4.205.35]) by mail1.shipmail.org (Postfix) with ESMTPSA id 94055360384; Wed, 18 Sep 2019 14:59:24 +0200 (CEST) From: =?utf-8?q?Thomas_Hellstr=C3=B6m_=28VMware=29?= To: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, linux-mm@kvack.org Subject: [PATCH 4/7] drm/vmwgfx: Implement an infrastructure for write-coherent resources Date: Wed, 18 Sep 2019 14:59:11 +0200 Message-Id: <20190918125914.38497-5-thomas_os@shipmail.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20190918125914.38497-1-thomas_os@shipmail.org> References: <20190918125914.38497-1-thomas_os@shipmail.org> MIME-Version: 1.0 X-Mailman-Original-DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=shipmail.org; s=mail; t=1568811564; bh=fwwR1uwCgnZ9/3IJ3lA6MBBx9Sz+7PEsg04qBIk0YoA=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=JWLvU3NAPU+MZMSCkPZFTLpOjRLZaC6WABziUwWzL45sj9D4sbbT9mcfxQocEDAd0 Ke3YVmB2a1fgXVzciXbUNrTdlnyZPT9DvAr9N4Ie0kTY9ejhxVzdCtGffBuinmBy0v bHjMFM+RUF369QCsK2bOD3dSV8Bf02hRL1X/I6Wg= X-Mailman-Original-Authentication-Results: ste-pvt-msa2.bahnhof.se; dkim=pass (1024-bit key; unprotected) header.d=shipmail.org header.i=@shipmail.org header.b=JWLvU3NA; dkim-atps=neutral X-Mailman-Original-Authentication-Results: ste-ftg-msa2.bahnhof.se (amavisd-new); dkim=pass (1024-bit key) header.d=shipmail.org X-BeenThere: dri-devel@lists.freedesktop.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Direct Rendering Infrastructure - Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Thomas Hellstrom , Michal Hocko , Rik van Riel , pv-drivers@vmware.com, Minchan Kim , Will Deacon , Matthew Wilcox , Christoph Hellwig , Peter Zijlstra , =?utf-8?b?SsOpcsO0bWUgR2xpc3Nl?= , linux-graphics-maintainer@vmware.com, Souptick Joarder , Huang Ying , Andrew Morton , Deepak Rawat , =?utf-8?q?Christian_K=C3=B6nig?= Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" From: Thomas Hellstrom This infrastructure will, for coherent resources, make sure that from the user-space point of view, data written by the CPU is immediately automatically available to the GPU at resource validation time. Cc: Andrew Morton Cc: Matthew Wilcox Cc: Will Deacon Cc: Peter Zijlstra Cc: Rik van Riel Cc: Minchan Kim Cc: Michal Hocko Cc: Huang Ying Cc: Souptick Joarder Cc: "Jérôme Glisse" Cc: "Christian König" Cc: Christoph Hellwig Signed-off-by: Thomas Hellstrom Reviewed-by: Deepak Rawat --- drivers/gpu/drm/vmwgfx/Kconfig | 1 + drivers/gpu/drm/vmwgfx/Makefile | 2 +- drivers/gpu/drm/vmwgfx/vmwgfx_bo.c | 5 +- drivers/gpu/drm/vmwgfx/vmwgfx_drv.h | 23 +- drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c | 1 - drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c | 417 ++++++++++++++++++ drivers/gpu/drm/vmwgfx/vmwgfx_resource.c | 57 +++ drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h | 11 + drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c | 15 +- drivers/gpu/drm/vmwgfx/vmwgfx_validation.c | 71 +++ drivers/gpu/drm/vmwgfx/vmwgfx_validation.h | 16 +- 11 files changed, 598 insertions(+), 21 deletions(-) create mode 100644 drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c diff --git a/drivers/gpu/drm/vmwgfx/Kconfig b/drivers/gpu/drm/vmwgfx/Kconfig index 6b28a326f8bb..d5fd81a521f6 100644 --- a/drivers/gpu/drm/vmwgfx/Kconfig +++ b/drivers/gpu/drm/vmwgfx/Kconfig @@ -8,6 +8,7 @@ config DRM_VMWGFX select FB_CFB_IMAGEBLIT select DRM_TTM select FB + select AS_DIRTY_HELPERS # Only needed for the transitional use of drm_crtc_init - can be removed # again once vmwgfx sets up the primary plane itself. select DRM_KMS_HELPER diff --git a/drivers/gpu/drm/vmwgfx/Makefile b/drivers/gpu/drm/vmwgfx/Makefile index 8841bd30e1e5..c877a21a0739 100644 --- a/drivers/gpu/drm/vmwgfx/Makefile +++ b/drivers/gpu/drm/vmwgfx/Makefile @@ -8,7 +8,7 @@ vmwgfx-y := vmwgfx_execbuf.o vmwgfx_gmr.o vmwgfx_kms.o vmwgfx_drv.o \ vmwgfx_cmdbuf_res.o vmwgfx_cmdbuf.o vmwgfx_stdu.o \ vmwgfx_cotable.o vmwgfx_so.o vmwgfx_binding.o vmwgfx_msg.o \ vmwgfx_simple_resource.o vmwgfx_va.o vmwgfx_blit.o \ - vmwgfx_validation.o \ + vmwgfx_validation.o vmwgfx_page_dirty.o \ ttm_object.o ttm_lock.o obj-$(CONFIG_DRM_VMWGFX) := vmwgfx.o diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c index aad8d8140259..869aeaec2f86 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c @@ -462,6 +462,7 @@ void vmw_bo_bo_free(struct ttm_buffer_object *bo) { struct vmw_buffer_object *vmw_bo = vmw_buffer_object(bo); + WARN_ON(vmw_bo->dirty); vmw_bo_unmap(vmw_bo); kfree(vmw_bo); } @@ -475,8 +476,10 @@ void vmw_bo_bo_free(struct ttm_buffer_object *bo) static void vmw_user_bo_destroy(struct ttm_buffer_object *bo) { struct vmw_user_buffer_object *vmw_user_bo = vmw_user_buffer_object(bo); + struct vmw_buffer_object *vbo = &vmw_user_bo->vbo; - vmw_bo_unmap(&vmw_user_bo->vbo); + WARN_ON(vbo->dirty); + vmw_bo_unmap(vbo); ttm_prime_object_kfree(vmw_user_bo, prime); } diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h index 5eb73ded8e07..7944dbbbdd72 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h @@ -105,6 +105,7 @@ struct vmw_fpriv { * @dx_query_ctx: DX context if this buffer object is used as a DX query MOB * @map: Kmap object for semi-persistent mappings * @res_prios: Eviction priority counts for attached resources + * @dirty: structure for user-space dirty-tracking */ struct vmw_buffer_object { struct ttm_buffer_object base; @@ -115,6 +116,7 @@ struct vmw_buffer_object { /* Protected by reservation */ struct ttm_bo_kmap_obj map; u32 res_prios[TTM_MAX_BO_PRIORITY]; + struct vmw_bo_dirty *dirty; }; /** @@ -145,7 +147,8 @@ struct vmw_res_func; * @res_dirty: Resource contains data not yet in the backup buffer. Protected * by resource reserved. * @backup_dirty: Backup buffer contains data not yet in the HW resource. - * Protecte by resource reserved. + * Protected by resource reserved. + * @coherent: Emulate coherency by tracking vm accesses. * @backup: The backup buffer if any. Protected by resource reserved. * @backup_offset: Offset into the backup buffer if any. Protected by resource * reserved. Note that only a few resource types can have a @backup_offset @@ -162,14 +165,16 @@ struct vmw_res_func; * @hw_destroy: Callback to destroy the resource on the device, as part of * resource destruction. */ +struct vmw_resource_dirty; struct vmw_resource { struct kref kref; struct vmw_private *dev_priv; int id; u32 used_prio; unsigned long backup_size; - bool res_dirty; - bool backup_dirty; + u32 res_dirty : 1; + u32 backup_dirty : 1; + u32 coherent : 1; struct vmw_buffer_object *backup; unsigned long backup_offset; unsigned long pin_count; @@ -177,6 +182,7 @@ struct vmw_resource { struct list_head lru_head; struct list_head mob_head; struct list_head binding_head; + struct vmw_resource_dirty *dirty; void (*res_free) (struct vmw_resource *res); void (*hw_destroy) (struct vmw_resource *res); }; @@ -716,6 +722,8 @@ extern void vmw_resource_evict_all(struct vmw_private *dev_priv); extern void vmw_resource_unbind_list(struct vmw_buffer_object *vbo); void vmw_resource_mob_attach(struct vmw_resource *res); void vmw_resource_mob_detach(struct vmw_resource *res); +void vmw_resource_dirty_update(struct vmw_resource *res, pgoff_t start, + pgoff_t end); /** * vmw_resource_mob_attached - Whether a resource currently has a mob attached @@ -1403,6 +1411,15 @@ int vmw_host_log(const char *log); #define VMW_DEBUG_USER(fmt, ...) \ DRM_DEBUG_DRIVER(fmt, ##__VA_ARGS__) +/* Resource dirtying - vmwgfx_page_dirty.c */ +void vmw_bo_dirty_scan(struct vmw_buffer_object *vbo); +int vmw_bo_dirty_add(struct vmw_buffer_object *vbo); +void vmw_bo_dirty_transfer_to_res(struct vmw_resource *res); +void vmw_bo_dirty_clear_res(struct vmw_resource *res); +void vmw_bo_dirty_release(struct vmw_buffer_object *vbo); +vm_fault_t vmw_bo_vm_fault(struct vm_fault *vmf); +vm_fault_t vmw_bo_vm_mkwrite(struct vm_fault *vmf); + /** * VMW_DEBUG_KMS - Debug output for kernel mode-setting * diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c index ff86d49dc5e8..934ad7c0c342 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c @@ -2560,7 +2560,6 @@ static int vmw_cmd_dx_check_subresource(struct vmw_private *dev_priv, offsetof(typeof(*cmd), sid)); cmd = container_of(header, typeof(*cmd), header); - return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface, VMW_RES_DIRTY_NONE, user_surface_converter, &cmd->sid, NULL); diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c b/drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c new file mode 100644 index 000000000000..be3302a8e309 --- /dev/null +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c @@ -0,0 +1,417 @@ +// SPDX-License-Identifier: GPL-2.0 OR MIT +/************************************************************************** + * + * Copyright 2019 VMware, Inc., Palo Alto, CA., USA + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the + * "Software"), to deal in the Software without restriction, including + * without limitation the rights to use, copy, modify, merge, publish, + * distribute, sub license, and/or sell copies of the Software, and to + * permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice (including the + * next paragraph) shall be included in all copies or substantial portions + * of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, + * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR + * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE + * USE OR OTHER DEALINGS IN THE SOFTWARE. + * + **************************************************************************/ +#include "vmwgfx_drv.h" + +/* + * Different methods for tracking dirty: + * VMW_BO_DIRTY_PAGETABLE - Scan the pagetable for hardware dirty bits + * VMW_BO_DIRTY_MKWRITE - Write-protect page table entries and record write- + * accesses in the VM mkwrite() callback + */ +enum vmw_bo_dirty_method { + VMW_BO_DIRTY_PAGETABLE, + VMW_BO_DIRTY_MKWRITE, +}; + +/* + * No dirtied pages at scan trigger a transition to the _MKWRITE method, + * similarly a certain percentage of dirty pages trigger a transition to + * the _PAGETABLE method. How many triggers should we wait for before + * changing method? + */ +#define VMW_DIRTY_NUM_CHANGE_TRIGGERS 2 + +/* Percentage to trigger a transition to the _PAGETABLE method */ +#define VMW_DIRTY_PERCENTAGE 10 + +/** + * struct vmw_bo_dirty - Dirty information for buffer objects + * @start: First currently dirty bit + * @end: Last currently dirty bit + 1 + * @method: The currently used dirty method + * @change_count: Number of consecutive method change triggers + * @ref_count: Reference count for this structure + * @bitmap_size: The size of the bitmap in bits. Typically equal to the + * nuber of pages in the bo. + * @size: The accounting size for this struct. + * @bitmap: A bitmap where each bit represents a page. A set bit means a + * dirty page. + */ +struct vmw_bo_dirty { + unsigned long start; + unsigned long end; + enum vmw_bo_dirty_method method; + unsigned int change_count; + unsigned int ref_count; + unsigned long bitmap_size; + size_t size; + unsigned long bitmap[0]; +}; + +/** + * vmw_bo_dirty_scan_pagetable - Perform a pagetable scan for dirty bits + * @vbo: The buffer object to scan + * + * Scans the pagetable for dirty bits. Clear those bits and modify the + * dirty structure with the results. This function may change the + * dirty-tracking method. + */ +static void vmw_bo_dirty_scan_pagetable(struct vmw_buffer_object *vbo) +{ + struct vmw_bo_dirty *dirty = vbo->dirty; + pgoff_t offset = drm_vma_node_start(&vbo->base.base.vma_node); + struct address_space *mapping = vbo->base.bdev->dev_mapping; + pgoff_t num_marked; + + num_marked = apply_as_clean(mapping, + offset, dirty->bitmap_size, + offset, &dirty->bitmap[0], + &dirty->start, &dirty->end); + if (num_marked == 0) + dirty->change_count++; + else + dirty->change_count = 0; + + if (dirty->change_count > VMW_DIRTY_NUM_CHANGE_TRIGGERS) { + dirty->change_count = 0; + dirty->method = VMW_BO_DIRTY_MKWRITE; + apply_as_wrprotect(mapping, + offset, dirty->bitmap_size); + apply_as_clean(mapping, + offset, dirty->bitmap_size, + offset, &dirty->bitmap[0], + &dirty->start, &dirty->end); + } +} + +/** + * vmw_bo_dirty_scan_mkwrite - Reset the mkwrite dirty-tracking method + * @vbo: The buffer object to scan + * + * Write-protect pages written to so that consecutive write accesses will + * trigger a call to mkwrite. + * + * This function may change the dirty-tracking method. + */ +static void vmw_bo_dirty_scan_mkwrite(struct vmw_buffer_object *vbo) +{ + struct vmw_bo_dirty *dirty = vbo->dirty; + unsigned long offset = drm_vma_node_start(&vbo->base.base.vma_node); + struct address_space *mapping = vbo->base.bdev->dev_mapping; + pgoff_t num_marked; + + if (dirty->end <= dirty->start) + return; + + num_marked = apply_as_wrprotect(vbo->base.bdev->dev_mapping, + dirty->start + offset, + dirty->end - dirty->start); + + if (100UL * num_marked / dirty->bitmap_size > + VMW_DIRTY_PERCENTAGE) { + dirty->change_count++; + } else { + dirty->change_count = 0; + } + + if (dirty->change_count > VMW_DIRTY_NUM_CHANGE_TRIGGERS) { + pgoff_t start = 0; + pgoff_t end = dirty->bitmap_size; + + dirty->method = VMW_BO_DIRTY_PAGETABLE; + apply_as_clean(mapping, offset, end, offset, &dirty->bitmap[0], + &start, &end); + bitmap_clear(&dirty->bitmap[0], 0, dirty->bitmap_size); + if (dirty->start < dirty->end) + bitmap_set(&dirty->bitmap[0], dirty->start, + dirty->end - dirty->start); + dirty->change_count = 0; + } +} + + +/** + * vmw_bo_dirty_scan - Scan for dirty pages and add them to the dirty + * tracking structure + * @vbo: The buffer object to scan + * + * This function may change the dirty tracking method. + */ +void vmw_bo_dirty_scan(struct vmw_buffer_object *vbo) +{ + struct vmw_bo_dirty *dirty = vbo->dirty; + + if (dirty->method == VMW_BO_DIRTY_PAGETABLE) + vmw_bo_dirty_scan_pagetable(vbo); + else + vmw_bo_dirty_scan_mkwrite(vbo); +} + +/** + * vmw_bo_dirty_add - Add a dirty-tracking user to a buffer object + * @vbo: The buffer object + * + * This function registers a dirty-tracking user to a buffer object. + * A user can be for example a resource or a vma in a special user-space + * mapping. + * + * Return: Zero on success, -ENOMEM on memory allocation failure. + */ +int vmw_bo_dirty_add(struct vmw_buffer_object *vbo) +{ + struct vmw_bo_dirty *dirty = vbo->dirty; + pgoff_t num_pages = vbo->base.num_pages; + size_t size, acc_size; + int ret; + static struct ttm_operation_ctx ctx = { + .interruptible = false, + .no_wait_gpu = false + }; + + if (dirty) { + dirty->ref_count++; + return 0; + } + + size = sizeof(*dirty) + BITS_TO_LONGS(num_pages) * sizeof(long); + acc_size = ttm_round_pot(size); + ret = ttm_mem_global_alloc(&ttm_mem_glob, acc_size, &ctx); + if (ret) { + VMW_DEBUG_USER("Out of graphics memory for buffer object " + "dirty tracker.\n"); + return ret; + } + dirty = kvzalloc(size, GFP_KERNEL); + if (!dirty) { + ret = -ENOMEM; + goto out_no_dirty; + } + + dirty->size = acc_size; + dirty->bitmap_size = num_pages; + dirty->start = dirty->bitmap_size; + dirty->end = 0; + dirty->ref_count = 1; + if (num_pages < PAGE_SIZE / sizeof(pte_t)) { + dirty->method = VMW_BO_DIRTY_PAGETABLE; + } else { + struct address_space *mapping = vbo->base.bdev->dev_mapping; + pgoff_t offset = drm_vma_node_start(&vbo->base.base.vma_node); + + dirty->method = VMW_BO_DIRTY_MKWRITE; + + /* Write-protect and then pick up already dirty bits */ + apply_as_wrprotect(mapping, offset, num_pages); + apply_as_clean(mapping, offset, num_pages, offset, + &dirty->bitmap[0], &dirty->start, &dirty->end); + } + + vbo->dirty = dirty; + + return 0; + +out_no_dirty: + ttm_mem_global_free(&ttm_mem_glob, acc_size); + return ret; +} + +/** + * vmw_bo_dirty_release - Release a dirty-tracking user from a buffer object + * @vbo: The buffer object + * + * This function releases a dirty-tracking user from a buffer object. + * If the reference count reaches zero, then the dirty-tracking object is + * freed and the pointer to it cleared. + * + * Return: Zero on success, -ENOMEM on memory allocation failure. + */ +void vmw_bo_dirty_release(struct vmw_buffer_object *vbo) +{ + struct vmw_bo_dirty *dirty = vbo->dirty; + + if (dirty && --dirty->ref_count == 0) { + size_t acc_size = dirty->size; + + kvfree(dirty); + ttm_mem_global_free(&ttm_mem_glob, acc_size); + vbo->dirty = NULL; + } +} + +/** + * vmw_bo_dirty_transfer_to_res - Pick up a resource's dirty region from + * its backing mob. + * @res: The resource + * + * This function will pick up all dirty ranges affecting the resource from + * it's backup mob, and call vmw_resource_dirty_update() once for each + * range. The transferred ranges will be cleared from the backing mob's + * dirty tracking. + */ +void vmw_bo_dirty_transfer_to_res(struct vmw_resource *res) +{ + struct vmw_buffer_object *vbo = res->backup; + struct vmw_bo_dirty *dirty = vbo->dirty; + pgoff_t start, cur, end; + unsigned long res_start = res->backup_offset; + unsigned long res_end = res->backup_offset + res->backup_size; + + WARN_ON_ONCE(res_start & ~PAGE_MASK); + res_start >>= PAGE_SHIFT; + res_end = DIV_ROUND_UP(res_end, PAGE_SIZE); + + if (res_start >= dirty->end || res_end <= dirty->start) + return; + + cur = max(res_start, dirty->start); + res_end = max(res_end, dirty->end); + while (cur < res_end) { + unsigned long num; + + start = find_next_bit(&dirty->bitmap[0], res_end, cur); + if (start >= res_end) + break; + + end = find_next_zero_bit(&dirty->bitmap[0], res_end, start + 1); + cur = end + 1; + num = end - start; + bitmap_clear(&dirty->bitmap[0], start, num); + vmw_resource_dirty_update(res, start, end); + } + + if (res_start <= dirty->start && res_end > dirty->start) + dirty->start = res_end; + if (res_start < dirty->end && res_end >= dirty->end) + dirty->end = res_start; +} + +/** + * vmw_bo_dirty_clear_res - Clear a resource's dirty region from + * its backing mob. + * @res: The resource + * + * This function will clear all dirty ranges affecting the resource from + * it's backup mob's dirty tracking. + */ +void vmw_bo_dirty_clear_res(struct vmw_resource *res) +{ + unsigned long res_start = res->backup_offset; + unsigned long res_end = res->backup_offset + res->backup_size; + struct vmw_buffer_object *vbo = res->backup; + struct vmw_bo_dirty *dirty = vbo->dirty; + + res_start >>= PAGE_SHIFT; + res_end = DIV_ROUND_UP(res_end, PAGE_SIZE); + + if (res_start >= dirty->end || res_end <= dirty->start) + return; + + res_start = max(res_start, dirty->start); + res_end = min(res_end, dirty->end); + bitmap_clear(&dirty->bitmap[0], res_start, res_end - res_start); + + if (res_start <= dirty->start && res_end > dirty->start) + dirty->start = res_end; + if (res_start < dirty->end && res_end >= dirty->end) + dirty->end = res_start; +} + +vm_fault_t vmw_bo_vm_mkwrite(struct vm_fault *vmf) +{ + struct vm_area_struct *vma = vmf->vma; + struct ttm_buffer_object *bo = (struct ttm_buffer_object *) + vma->vm_private_data; + vm_fault_t ret; + unsigned long page_offset; + unsigned int save_flags; + struct vmw_buffer_object *vbo = + container_of(bo, typeof(*vbo), base); + + /* + * mkwrite() doesn't handle the VM_FAULT_RETRY return value correctly. + * So make sure the TTM helpers are aware. + */ + save_flags = vmf->flags; + vmf->flags &= ~FAULT_FLAG_ALLOW_RETRY; + ret = ttm_bo_vm_reserve(bo, vmf); + vmf->flags = save_flags; + if (ret) + return ret; + + page_offset = vmf->pgoff - drm_vma_node_start(&bo->base.vma_node); + if (unlikely(page_offset >= bo->num_pages)) { + ret = VM_FAULT_SIGBUS; + goto out_unlock; + } + + if (vbo->dirty && vbo->dirty->method == VMW_BO_DIRTY_MKWRITE && + !test_bit(page_offset, &vbo->dirty->bitmap[0])) { + struct vmw_bo_dirty *dirty = vbo->dirty; + + __set_bit(page_offset, &dirty->bitmap[0]); + dirty->start = min(dirty->start, page_offset); + dirty->end = max(dirty->end, page_offset + 1); + } + +out_unlock: + dma_resv_unlock(bo->base.resv); + return ret; +} + +vm_fault_t vmw_bo_vm_fault(struct vm_fault *vmf) +{ + struct vm_area_struct *vma = vmf->vma; + struct ttm_buffer_object *bo = (struct ttm_buffer_object *) + vma->vm_private_data; + struct vmw_buffer_object *vbo = + container_of(bo, struct vmw_buffer_object, base); + pgoff_t num_prefault; + pgprot_t prot; + vm_fault_t ret; + + ret = ttm_bo_vm_reserve(bo, vmf); + if (ret) + return ret; + + /* + * This will cause mkwrite() to be called for each pte on + * write-enable vmas. + */ + if (vbo->dirty && vbo->dirty->method == VMW_BO_DIRTY_MKWRITE) + prot = vma->vm_page_prot; + else + prot = vm_get_page_prot(vma->vm_flags); + + num_prefault = (vma->vm_flags & VM_RAND_READ) ? 0 : + TTM_BO_VM_NUM_PREFAULT; + ret = ttm_bo_vm_fault_reserved(vmf, prot, num_prefault); + if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) + return ret; + + dma_resv_unlock(bo->base.resv); + return ret; +} diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c index 5581a7826b4c..e4c97a4cf2ff 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c @@ -119,6 +119,10 @@ static void vmw_resource_release(struct kref *kref) } res->backup_dirty = false; vmw_resource_mob_detach(res); + if (res->dirty) + res->func->dirty_free(res); + if (res->coherent) + vmw_bo_dirty_release(res->backup); ttm_bo_unreserve(bo); vmw_bo_unreference(&res->backup); } @@ -208,7 +212,9 @@ int vmw_resource_init(struct vmw_private *dev_priv, struct vmw_resource *res, res->backup_offset = 0; res->backup_dirty = false; res->res_dirty = false; + res->coherent = false; res->used_prio = 3; + res->dirty = NULL; if (delay_id) return 0; else @@ -395,6 +401,30 @@ static int vmw_resource_do_validate(struct vmw_resource *res, vmw_resource_mob_attach(res); } + /* + * Handle the case where the backup mob is marked coherent but + * the resource isn't. + */ + if (func->dirty_alloc && vmw_resource_mob_attached(res) && + !res->coherent) { + if (res->backup->dirty && !res->dirty) { + ret = func->dirty_alloc(res); + if (ret) + return ret; + } else if (!res->backup->dirty && res->dirty) { + func->dirty_free(res); + } + } + + /* + * Transfer the dirty regions to the resource and update + * the resource. + */ + if (res->dirty) { + vmw_bo_dirty_transfer_to_res(res); + return func->dirty_sync(res); + } + return 0; out_bind_failed: @@ -433,16 +463,28 @@ void vmw_resource_unreserve(struct vmw_resource *res, if (switch_backup && new_backup != res->backup) { if (res->backup) { vmw_resource_mob_detach(res); + if (res->coherent) + vmw_bo_dirty_release(res->backup); vmw_bo_unreference(&res->backup); } if (new_backup) { res->backup = vmw_bo_reference(new_backup); + + /* + * The validation code should already have added a + * dirty tracker here. + */ + WARN_ON(res->coherent && !new_backup->dirty); + vmw_resource_mob_attach(res); } else { res->backup = NULL; } + } else if (switch_backup && res->coherent) { + vmw_bo_dirty_release(res->backup); } + if (switch_backup) res->backup_offset = new_backup_offset; @@ -1008,3 +1050,18 @@ enum vmw_res_type vmw_res_type(const struct vmw_resource *res) { return res->func->res_type; } + +/** + * vmw_resource_update_dirty - Update a resource's dirty tracker with a + * sequential range of touched backing store memory. + * @res: The resource. + * @start: The first page touched. + * @end: The last page touched + 1. + */ +void vmw_resource_dirty_update(struct vmw_resource *res, pgoff_t start, + pgoff_t end) +{ + if (res->dirty) + res->func->dirty_range_add(res, start << PAGE_SHIFT, + end << PAGE_SHIFT); +} diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h index 984e588c62ca..c85144286cfe 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h @@ -71,6 +71,12 @@ struct vmw_user_resource_conv { * @commit_notify: If the resource is a command buffer managed resource, * callback to notify that a define or remove command * has been committed to the device. + * @dirty_alloc: Allocate a dirty tracker. NULL if dirty-tracking is not + * supported. + * @dirty_free: Free the dirty tracker. + * @dirty_sync: Upload the dirty mob contents to the resource. + * @dirty_add_range: Add a sequential dirty range to the resource + * dirty tracker. */ struct vmw_res_func { enum vmw_res_type res_type; @@ -90,6 +96,11 @@ struct vmw_res_func { struct ttm_validate_buffer *val_buf); void (*commit_notify)(struct vmw_resource *res, enum vmw_cmdbuf_res_state state); + int (*dirty_alloc)(struct vmw_resource *res); + void (*dirty_free)(struct vmw_resource *res); + int (*dirty_sync)(struct vmw_resource *res); + void (*dirty_range_add)(struct vmw_resource *res, size_t start, + size_t end); }; /** diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c index 5a7b8bb420de..ce288756531b 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c @@ -29,10 +29,23 @@ int vmw_mmap(struct file *filp, struct vm_area_struct *vma) { + static const struct vm_operations_struct vmw_vm_ops = { + .pfn_mkwrite = vmw_bo_vm_mkwrite, + .page_mkwrite = vmw_bo_vm_mkwrite, + .fault = vmw_bo_vm_fault, + .open = ttm_bo_vm_open, + .close = ttm_bo_vm_close + }; struct drm_file *file_priv = filp->private_data; struct vmw_private *dev_priv = vmw_priv(file_priv->minor->dev); + int ret = ttm_bo_mmap(filp, vma, &dev_priv->bdev); - return ttm_bo_mmap(filp, vma, &dev_priv->bdev); + if (ret) + return ret; + + vma->vm_ops = &vmw_vm_ops; + + return 0; } /* struct vmw_validation_mem callback */ diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c index f611b2290a1b..71349a7bae90 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c @@ -33,6 +33,8 @@ * struct vmw_validation_bo_node - Buffer object validation metadata. * @base: Metadata used for TTM reservation- and validation. * @hash: A hash entry used for the duplicate detection hash table. + * @coherent_count: If switching backup buffers, number of new coherent + * resources that will have this buffer as a backup buffer. * @as_mob: Validate as mob. * @cpu_blit: Validate for cpu blit access. * @@ -42,6 +44,7 @@ struct vmw_validation_bo_node { struct ttm_validate_buffer base; struct drm_hash_item hash; + unsigned int coherent_count; u32 as_mob : 1; u32 cpu_blit : 1; }; @@ -459,6 +462,19 @@ int vmw_validation_res_reserve(struct vmw_validation_context *ctx, if (ret) goto out_unreserve; } + + if (val->switching_backup && val->new_backup && + res->coherent) { + struct vmw_validation_bo_node *bo_node = + vmw_validation_find_bo_dup(ctx, + val->new_backup); + + if (WARN_ON(!bo_node)) { + ret = -EINVAL; + goto out_unreserve; + } + bo_node->coherent_count++; + } } return 0; @@ -562,6 +578,9 @@ int vmw_validation_bo_validate(struct vmw_validation_context *ctx, bool intr) int ret; list_for_each_entry(entry, &ctx->bo_list, base.head) { + struct vmw_buffer_object *vbo = + container_of(entry->base.bo, typeof(*vbo), base); + if (entry->cpu_blit) { struct ttm_operation_ctx ctx = { .interruptible = intr, @@ -576,6 +595,27 @@ int vmw_validation_bo_validate(struct vmw_validation_context *ctx, bool intr) } if (ret) return ret; + + /* + * Rather than having the resource code allocating the bo + * dirty tracker in resource_unreserve() where we can't fail, + * Do it here when validating the buffer object. + */ + if (entry->coherent_count) { + unsigned int coherent_count = entry->coherent_count; + + while (coherent_count) { + ret = vmw_bo_dirty_add(vbo); + if (ret) + return ret; + + coherent_count--; + } + entry->coherent_count -= coherent_count; + } + + if (vbo->dirty) + vmw_bo_dirty_scan(vbo); } return 0; } @@ -828,3 +868,34 @@ int vmw_validation_preload_res(struct vmw_validation_context *ctx, ctx->mem_size_left += size; return 0; } + +/** + * vmw_validation_bo_backoff - Unreserve buffer objects registered with a + * validation context + * @ctx: The validation context + * + * This function unreserves the buffer objects previously reserved using + * vmw_validation_bo_reserve. It's typically used as part of an error path + */ +void vmw_validation_bo_backoff(struct vmw_validation_context *ctx) +{ + struct vmw_validation_bo_node *entry; + + /* + * Switching coherent resource backup buffers failed. + * Release corresponding buffer object dirty trackers. + */ + list_for_each_entry(entry, &ctx->bo_list, base.head) { + if (entry->coherent_count) { + unsigned int coherent_count = entry->coherent_count; + struct vmw_buffer_object *vbo = + container_of(entry->base.bo, typeof(*vbo), + base); + + while (coherent_count--) + vmw_bo_dirty_release(vbo); + } + } + + ttm_eu_backoff_reservation(&ctx->ticket, &ctx->bo_list); +} diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.h b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.h index 0e063743dd86..4cee3f732588 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.h @@ -173,20 +173,6 @@ vmw_validation_bo_reserve(struct vmw_validation_context *ctx, NULL, true); } -/** - * vmw_validation_bo_backoff - Unreserve buffer objects registered with a - * validation context - * @ctx: The validation context - * - * This function unreserves the buffer objects previously reserved using - * vmw_validation_bo_reserve. It's typically used as part of an error path - */ -static inline void -vmw_validation_bo_backoff(struct vmw_validation_context *ctx) -{ - ttm_eu_backoff_reservation(&ctx->ticket, &ctx->bo_list); -} - /** * vmw_validation_bo_fence - Unreserve and fence buffer objects registered * with a validation context @@ -269,4 +255,6 @@ int vmw_validation_preload_res(struct vmw_validation_context *ctx, unsigned int size); void vmw_validation_res_set_dirty(struct vmw_validation_context *ctx, void *val_private, u32 dirty); +void vmw_validation_bo_backoff(struct vmw_validation_context *ctx); + #endif From patchwork Wed Sep 18 12:59:12 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: =?utf-8?q?Thomas_Hellstr=C3=B6m_=28Intel=29?= X-Patchwork-Id: 11150313 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id AC0BC912 for ; Wed, 18 Sep 2019 12:59:41 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 941AD21924 for ; Wed, 18 Sep 2019 12:59:41 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 941AD21924 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=shipmail.org Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=dri-devel-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id BDA3D6EF34; Wed, 18 Sep 2019 12:59:37 +0000 (UTC) X-Original-To: dri-devel@lists.freedesktop.org Delivered-To: dri-devel@lists.freedesktop.org Received: from ste-pvt-msa2.bahnhof.se (ste-pvt-msa2.bahnhof.se [213.80.101.71]) by gabe.freedesktop.org (Postfix) with ESMTPS id 2C89F6EF31 for ; Wed, 18 Sep 2019 12:59:34 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by ste-pvt-msa2.bahnhof.se (Postfix) with ESMTP id 28AB33F76F; Wed, 18 Sep 2019 14:59:27 +0200 (CEST) X-Virus-Scanned: Debian amavisd-new at bahnhof.se X-Spam-Flag: NO X-Spam-Score: -2.099 X-Spam-Level: X-Spam-Status: No, score=-2.099 tagged_above=-999 required=6.31 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no Received: from ste-pvt-msa2.bahnhof.se ([127.0.0.1]) by localhost (ste-ftg-msa2.bahnhof.se [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Fivy1EwBAv0d; Wed, 18 Sep 2019 14:59:26 +0200 (CEST) Received: from mail1.shipmail.org (h-205-35.A357.priv.bahnhof.se [155.4.205.35]) (Authenticated sender: mb878879) by ste-pvt-msa2.bahnhof.se (Postfix) with ESMTPA id 058013F75E; Wed, 18 Sep 2019 14:59:26 +0200 (CEST) Received: from localhost.localdomain.localdomain (h-205-35.A357.priv.bahnhof.se [155.4.205.35]) by mail1.shipmail.org (Postfix) with ESMTPSA id CC5D7360596; Wed, 18 Sep 2019 14:59:24 +0200 (CEST) From: =?utf-8?q?Thomas_Hellstr=C3=B6m_=28VMware=29?= To: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, linux-mm@kvack.org Subject: [PATCH 5/7] drm/vmwgfx: Use an RBtree instead of linked list for MOB resources Date: Wed, 18 Sep 2019 14:59:12 +0200 Message-Id: <20190918125914.38497-6-thomas_os@shipmail.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20190918125914.38497-1-thomas_os@shipmail.org> References: <20190918125914.38497-1-thomas_os@shipmail.org> MIME-Version: 1.0 X-Mailman-Original-DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=shipmail.org; s=mail; t=1568811564; bh=SNbcWxMXta8qaUE8qMDRNvV7AMAIqFn6HCBuIiXlnnE=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=IzRdp8+v7fA4LMekLnj0MXiJtJBLq7XeZlx+7y0zwbTlL2u4xunODnSG8tZPd/m3u 6TuKjbIE7fxf3euoq6e383hZPhc8vxvm02AJXyoiku97K7G9BOIUCPWvg4t3/VQuOW SbI1JeyHO8wCmE0jTOoHT251CipJNnDcW2uxYOQU= X-Mailman-Original-Authentication-Results: ste-pvt-msa2.bahnhof.se; dkim=pass (1024-bit key; unprotected) header.d=shipmail.org header.i=@shipmail.org header.b=IzRdp8+v; dkim-atps=neutral X-Mailman-Original-Authentication-Results: ste-ftg-msa2.bahnhof.se (amavisd-new); dkim=pass (1024-bit key) header.d=shipmail.org X-BeenThere: dri-devel@lists.freedesktop.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Direct Rendering Infrastructure - Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Thomas Hellstrom , Michal Hocko , Rik van Riel , pv-drivers@vmware.com, Minchan Kim , Will Deacon , Matthew Wilcox , Christoph Hellwig , Peter Zijlstra , =?utf-8?b?SsOpcsO0bWUgR2xpc3Nl?= , linux-graphics-maintainer@vmware.com, Souptick Joarder , Huang Ying , Andrew Morton , Deepak Rawat , =?utf-8?q?Christian_K=C3=B6nig?= Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" From: Thomas Hellstrom With emulated coherent memory we need to be able to quickly look up a resource from the MOB offset. Instead of traversing a linked list with O(n) worst case, use an RBtree with O(log n) worst case complexity. Cc: Andrew Morton Cc: Matthew Wilcox Cc: Will Deacon Cc: Peter Zijlstra Cc: Rik van Riel Cc: Minchan Kim Cc: Michal Hocko Cc: Huang Ying Cc: Souptick Joarder Cc: "Jérôme Glisse" Cc: "Christian König" Cc: Christoph Hellwig Signed-off-by: Thomas Hellstrom Reviewed-by: Deepak Rawat --- drivers/gpu/drm/vmwgfx/vmwgfx_bo.c | 5 ++-- drivers/gpu/drm/vmwgfx/vmwgfx_drv.h | 10 +++---- drivers/gpu/drm/vmwgfx/vmwgfx_resource.c | 33 +++++++++++++++++------- 3 files changed, 32 insertions(+), 16 deletions(-) diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c index 869aeaec2f86..18e4b329e563 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c @@ -463,6 +463,7 @@ void vmw_bo_bo_free(struct ttm_buffer_object *bo) struct vmw_buffer_object *vmw_bo = vmw_buffer_object(bo); WARN_ON(vmw_bo->dirty); + WARN_ON(!RB_EMPTY_ROOT(&vmw_bo->res_tree)); vmw_bo_unmap(vmw_bo); kfree(vmw_bo); } @@ -479,6 +480,7 @@ static void vmw_user_bo_destroy(struct ttm_buffer_object *bo) struct vmw_buffer_object *vbo = &vmw_user_bo->vbo; WARN_ON(vbo->dirty); + WARN_ON(!RB_EMPTY_ROOT(&vbo->res_tree)); vmw_bo_unmap(vbo); ttm_prime_object_kfree(vmw_user_bo, prime); } @@ -514,8 +516,7 @@ int vmw_bo_init(struct vmw_private *dev_priv, memset(vmw_bo, 0, sizeof(*vmw_bo)); BUILD_BUG_ON(TTM_MAX_BO_PRIORITY <= 3); vmw_bo->base.priority = 3; - - INIT_LIST_HEAD(&vmw_bo->res_list); + vmw_bo->res_tree = RB_ROOT; ret = ttm_bo_init(bdev, &vmw_bo->base, size, ttm_bo_type_device, placement, diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h index 7944dbbbdd72..53f8522ae032 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h @@ -100,7 +100,7 @@ struct vmw_fpriv { /** * struct vmw_buffer_object - TTM buffer object with vmwgfx additions * @base: The TTM buffer object - * @res_list: List of resources using this buffer object as a backing MOB + * @res_tree: RB tree of resources using this buffer object as a backing MOB * @pin_count: pin depth * @dx_query_ctx: DX context if this buffer object is used as a DX query MOB * @map: Kmap object for semi-persistent mappings @@ -109,7 +109,7 @@ struct vmw_fpriv { */ struct vmw_buffer_object { struct ttm_buffer_object base; - struct list_head res_list; + struct rb_root res_tree; s32 pin_count; /* Not ref-counted. Protected by binding_mutex */ struct vmw_resource *dx_query_ctx; @@ -157,8 +157,8 @@ struct vmw_res_func; * pin-count greater than zero. It is not on the resource LRU lists and its * backup buffer is pinned. Hence it can't be evicted. * @func: Method vtable for this resource. Immutable. + * @mob_node; Node for the MOB backup rbtree. Protected by @backup reserved. * @lru_head: List head for the LRU list. Protected by @dev_priv::resource_lock. - * @mob_head: List head for the MOB backup list. Protected by @backup reserved. * @binding_head: List head for the context binding list. Protected by * the @dev_priv::binding_mutex * @res_free: The resource destructor. @@ -179,8 +179,8 @@ struct vmw_resource { unsigned long backup_offset; unsigned long pin_count; const struct vmw_res_func *func; + struct rb_node mob_node; struct list_head lru_head; - struct list_head mob_head; struct list_head binding_head; struct vmw_resource_dirty *dirty; void (*res_free) (struct vmw_resource *res); @@ -733,7 +733,7 @@ void vmw_resource_dirty_update(struct vmw_resource *res, pgoff_t start, */ static inline bool vmw_resource_mob_attached(const struct vmw_resource *res) { - return !list_empty(&res->mob_head); + return !RB_EMPTY_NODE(&res->mob_node); } /** diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c index e4c97a4cf2ff..328ad46076ff 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c @@ -40,11 +40,24 @@ void vmw_resource_mob_attach(struct vmw_resource *res) { struct vmw_buffer_object *backup = res->backup; + struct rb_node **new = &backup->res_tree.rb_node, *parent = NULL; dma_resv_assert_held(res->backup->base.base.resv); res->used_prio = (res->res_dirty) ? res->func->dirty_prio : res->func->prio; - list_add_tail(&res->mob_head, &backup->res_list); + + while (*new) { + struct vmw_resource *this = + container_of(*new, struct vmw_resource, mob_node); + + parent = *new; + new = (res->backup_offset < this->backup_offset) ? + &((*new)->rb_left) : &((*new)->rb_right); + } + + rb_link_node(&res->mob_node, parent, new); + rb_insert_color(&res->mob_node, &backup->res_tree); + vmw_bo_prio_add(backup, res->used_prio); } @@ -58,7 +71,8 @@ void vmw_resource_mob_detach(struct vmw_resource *res) dma_resv_assert_held(backup->base.base.resv); if (vmw_resource_mob_attached(res)) { - list_del_init(&res->mob_head); + rb_erase(&res->mob_node, &backup->res_tree); + RB_CLEAR_NODE(&res->mob_node); vmw_bo_prio_del(backup, res->used_prio); } } @@ -204,8 +218,8 @@ int vmw_resource_init(struct vmw_private *dev_priv, struct vmw_resource *res, res->res_free = res_free; res->dev_priv = dev_priv; res->func = func; + RB_CLEAR_NODE(&res->mob_node); INIT_LIST_HEAD(&res->lru_head); - INIT_LIST_HEAD(&res->mob_head); INIT_LIST_HEAD(&res->binding_head); res->id = -1; res->backup = NULL; @@ -754,19 +768,20 @@ int vmw_resource_validate(struct vmw_resource *res, bool intr) */ void vmw_resource_unbind_list(struct vmw_buffer_object *vbo) { - - struct vmw_resource *res, *next; struct ttm_validate_buffer val_buf = { .bo = &vbo->base, .num_shared = 0 }; dma_resv_assert_held(vbo->base.base.resv); - list_for_each_entry_safe(res, next, &vbo->res_list, mob_head) { - if (!res->func->unbind) - continue; + while (!RB_EMPTY_ROOT(&vbo->res_tree)) { + struct rb_node *node = vbo->res_tree.rb_node; + struct vmw_resource *res = + container_of(node, struct vmw_resource, mob_node); + + if (!WARN_ON_ONCE(!res->func->unbind)) + (void) res->func->unbind(res, res->res_dirty, &val_buf); - (void) res->func->unbind(res, res->res_dirty, &val_buf); res->backup_dirty = true; res->res_dirty = false; vmw_resource_mob_detach(res); From patchwork Wed Sep 18 12:59:13 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: =?utf-8?q?Thomas_Hellstr=C3=B6m_=28Intel=29?= X-Patchwork-Id: 11150317 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id A6DC814DB for ; Wed, 18 Sep 2019 12:59:43 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 8E07421928 for ; Wed, 18 Sep 2019 12:59:43 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 8E07421928 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=shipmail.org Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=dri-devel-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 0B1836EF38; Wed, 18 Sep 2019 12:59:38 +0000 (UTC) X-Original-To: dri-devel@lists.freedesktop.org Delivered-To: dri-devel@lists.freedesktop.org Received: from ste-pvt-msa1.bahnhof.se (ste-pvt-msa1.bahnhof.se [213.80.101.70]) by gabe.freedesktop.org (Postfix) with ESMTPS id 2043F6EF31 for ; Wed, 18 Sep 2019 12:59:31 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by ste-pvt-msa1.bahnhof.se (Postfix) with ESMTP id 30B4E3F6DF; Wed, 18 Sep 2019 14:59:29 +0200 (CEST) X-Virus-Scanned: Debian amavisd-new at bahnhof.se X-Spam-Flag: NO X-Spam-Score: -2.099 X-Spam-Level: X-Spam-Status: No, score=-2.099 tagged_above=-999 required=6.31 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no Received: from ste-pvt-msa1.bahnhof.se ([127.0.0.1]) by localhost (ste-pvt-msa1.bahnhof.se [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id KD6nOvhenirs; Wed, 18 Sep 2019 14:59:27 +0200 (CEST) Received: from mail1.shipmail.org (h-205-35.A357.priv.bahnhof.se [155.4.205.35]) (Authenticated sender: mb878879) by ste-pvt-msa1.bahnhof.se (Postfix) with ESMTPA id 0C5EA3F6DC; Wed, 18 Sep 2019 14:59:26 +0200 (CEST) Received: from localhost.localdomain.localdomain (h-205-35.A357.priv.bahnhof.se [155.4.205.35]) by mail1.shipmail.org (Postfix) with ESMTPSA id 033C33605D3; Wed, 18 Sep 2019 14:59:25 +0200 (CEST) From: =?utf-8?q?Thomas_Hellstr=C3=B6m_=28VMware=29?= To: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, linux-mm@kvack.org Subject: [PATCH 6/7] drm/vmwgfx: Implement an infrastructure for read-coherent resources Date: Wed, 18 Sep 2019 14:59:13 +0200 Message-Id: <20190918125914.38497-7-thomas_os@shipmail.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20190918125914.38497-1-thomas_os@shipmail.org> References: <20190918125914.38497-1-thomas_os@shipmail.org> MIME-Version: 1.0 X-Mailman-Original-DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=shipmail.org; s=mail; t=1568811565; bh=egdMJrcK560ENyHe9ydAMcesRhcZhIUfGjqOtDGjqjk=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=ZtbPAoCfnYi1AVH1EMWEoiJK/3Sh4kzaQsLiHfC6HMrwUgJsaHi7HRr5yVMTT8Esr 34mLvDf0j/OO5KCdOZ+XC2ttXp3SoQ8tIlKvcjpiZQ1uLUAyAIJHHT7SqJKWLUQPQV fkwmgA1RJ43FZlBH9ul5kjKxFVsgglA/pXYNFQdA= X-Mailman-Original-Authentication-Results: ste-pvt-msa1.bahnhof.se; dkim=pass (1024-bit key; unprotected) header.d=shipmail.org header.i=@shipmail.org header.b=ZtbPAoCf; dkim-atps=neutral X-BeenThere: dri-devel@lists.freedesktop.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Direct Rendering Infrastructure - Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Thomas Hellstrom , Michal Hocko , Rik van Riel , pv-drivers@vmware.com, Minchan Kim , Will Deacon , Matthew Wilcox , Christoph Hellwig , Peter Zijlstra , =?utf-8?b?SsOpcsO0bWUgR2xpc3Nl?= , linux-graphics-maintainer@vmware.com, Souptick Joarder , Huang Ying , Andrew Morton , Deepak Rawat , =?utf-8?q?Christian_K=C3=B6nig?= Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" From: Thomas Hellstrom Similar to write-coherent resources, make sure that from the user-space point of view, GPU rendered contents is automatically available for reading by the CPU. Cc: Andrew Morton Cc: Matthew Wilcox Cc: Will Deacon Cc: Peter Zijlstra Cc: Rik van Riel Cc: Minchan Kim Cc: Michal Hocko Cc: Huang Ying Cc: Souptick Joarder Cc: "Jérôme Glisse" Cc: "Christian König" Cc: Christoph Hellwig Signed-off-by: Thomas Hellstrom Reviewed-by: Deepak Rawat --- drivers/gpu/drm/vmwgfx/vmwgfx_drv.h | 7 +- drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c | 75 ++++++++++++- drivers/gpu/drm/vmwgfx/vmwgfx_resource.c | 103 +++++++++++++++++- drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h | 2 + drivers/gpu/drm/vmwgfx/vmwgfx_validation.c | 3 +- 5 files changed, 179 insertions(+), 11 deletions(-) diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h index 53f8522ae032..729a2e93acf1 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h @@ -680,7 +680,8 @@ extern void vmw_resource_unreference(struct vmw_resource **p_res); extern struct vmw_resource *vmw_resource_reference(struct vmw_resource *res); extern struct vmw_resource * vmw_resource_reference_unless_doomed(struct vmw_resource *res); -extern int vmw_resource_validate(struct vmw_resource *res, bool intr); +extern int vmw_resource_validate(struct vmw_resource *res, bool intr, + bool dirtying); extern int vmw_resource_reserve(struct vmw_resource *res, bool interruptible, bool no_backup); extern bool vmw_resource_needs_backup(const struct vmw_resource *res); @@ -724,6 +725,8 @@ void vmw_resource_mob_attach(struct vmw_resource *res); void vmw_resource_mob_detach(struct vmw_resource *res); void vmw_resource_dirty_update(struct vmw_resource *res, pgoff_t start, pgoff_t end); +int vmw_resources_clean(struct vmw_buffer_object *vbo, pgoff_t start, + pgoff_t end, pgoff_t *num_prefault); /** * vmw_resource_mob_attached - Whether a resource currently has a mob attached @@ -1417,6 +1420,8 @@ int vmw_bo_dirty_add(struct vmw_buffer_object *vbo); void vmw_bo_dirty_transfer_to_res(struct vmw_resource *res); void vmw_bo_dirty_clear_res(struct vmw_resource *res); void vmw_bo_dirty_release(struct vmw_buffer_object *vbo); +void vmw_bo_dirty_unmap(struct vmw_buffer_object *vbo, + pgoff_t start, pgoff_t end); vm_fault_t vmw_bo_vm_fault(struct vm_fault *vmf); vm_fault_t vmw_bo_vm_mkwrite(struct vm_fault *vmf); diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c b/drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c index be3302a8e309..1914d34c183a 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c @@ -153,7 +153,6 @@ static void vmw_bo_dirty_scan_mkwrite(struct vmw_buffer_object *vbo) } } - /** * vmw_bo_dirty_scan - Scan for dirty pages and add them to the dirty * tracking structure @@ -171,6 +170,51 @@ void vmw_bo_dirty_scan(struct vmw_buffer_object *vbo) vmw_bo_dirty_scan_mkwrite(vbo); } +/** + * vmw_bo_dirty_pre_unmap - write-protect and pick up dirty pages before + * an unmap_mapping_range operation. + * @vbo: The buffer object, + * @start: First page of the range within the buffer object. + * @end: Last page of the range within the buffer object + 1. + * + * If we're using the _PAGETABLE scan method, we may leak dirty pages + * when calling unmap_mapping_range(). This function makes sure we pick + * up all dirty pages. + */ +static void vmw_bo_dirty_pre_unmap(struct vmw_buffer_object *vbo, + pgoff_t start, pgoff_t end) +{ + struct vmw_bo_dirty *dirty = vbo->dirty; + unsigned long offset = drm_vma_node_start(&vbo->base.base.vma_node); + struct address_space *mapping = vbo->base.bdev->dev_mapping; + + if (dirty->method != VMW_BO_DIRTY_PAGETABLE || start >= end) + return; + + apply_as_wrprotect(mapping, start + offset, end - start); + apply_as_clean(mapping, start + offset, end - start, offset, + &dirty->bitmap[0], &dirty->start, &dirty->end); +} + +/** + * vmw_bo_dirty_unmap - Clear all ptes pointing to a range within a bo + * @vbo: The buffer object, + * @start: First page of the range within the buffer object. + * @end: Last page of the range within the buffer object + 1. + * + * This is similar to ttm_bo_unmap_virtual_locked() except it takes a subrange. + */ +void vmw_bo_dirty_unmap(struct vmw_buffer_object *vbo, + pgoff_t start, pgoff_t end) +{ + unsigned long offset = drm_vma_node_start(&vbo->base.base.vma_node); + struct address_space *mapping = vbo->base.bdev->dev_mapping; + + vmw_bo_dirty_pre_unmap(vbo, start, end); + unmap_shared_mapping_range(mapping, (offset + start) << PAGE_SHIFT, + (loff_t) (end - start) << PAGE_SHIFT); +} + /** * vmw_bo_dirty_add - Add a dirty-tracking user to a buffer object * @vbo: The buffer object @@ -397,21 +441,42 @@ vm_fault_t vmw_bo_vm_fault(struct vm_fault *vmf) if (ret) return ret; + num_prefault = (vma->vm_flags & VM_RAND_READ) ? 1 : + TTM_BO_VM_NUM_PREFAULT; + + if (vbo->dirty) { + pgoff_t allowed_prefault; + unsigned long page_offset; + + page_offset = vmf->pgoff - + drm_vma_node_start(&bo->base.vma_node); + if (page_offset >= bo->num_pages || + vmw_resources_clean(vbo, page_offset, + page_offset + PAGE_SIZE, + &allowed_prefault)) { + ret = VM_FAULT_SIGBUS; + goto out_unlock; + } + + num_prefault = min(num_prefault, allowed_prefault); + } + /* - * This will cause mkwrite() to be called for each pte on - * write-enable vmas. + * If we don't track dirty using the MKWRITE method, make sure + * sure the page protection is write-enabled so we don't get + * a lot of unnecessary write faults. */ if (vbo->dirty && vbo->dirty->method == VMW_BO_DIRTY_MKWRITE) prot = vma->vm_page_prot; else prot = vm_get_page_prot(vma->vm_flags); - num_prefault = (vma->vm_flags & VM_RAND_READ) ? 0 : - TTM_BO_VM_NUM_PREFAULT; ret = ttm_bo_vm_fault_reserved(vmf, prot, num_prefault); if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) return ret; +out_unlock: dma_resv_unlock(bo->base.resv); + return ret; } diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c index 328ad46076ff..c76faf33972e 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c @@ -393,7 +393,8 @@ static int vmw_resource_buf_alloc(struct vmw_resource *res, * should be retried once resources have been freed up. */ static int vmw_resource_do_validate(struct vmw_resource *res, - struct ttm_validate_buffer *val_buf) + struct ttm_validate_buffer *val_buf, + bool dirtying) { int ret = 0; const struct vmw_res_func *func = res->func; @@ -435,6 +436,15 @@ static int vmw_resource_do_validate(struct vmw_resource *res, * the resource. */ if (res->dirty) { + if (dirtying && !res->res_dirty) { + pgoff_t start = res->backup_offset >> PAGE_SHIFT; + pgoff_t end = __KERNEL_DIV_ROUND_UP + (res->backup_offset + res->backup_size, + PAGE_SIZE); + + vmw_bo_dirty_unmap(res->backup, start, end); + } + vmw_bo_dirty_transfer_to_res(res); return func->dirty_sync(res); } @@ -679,6 +689,7 @@ static int vmw_resource_do_evict(struct ww_acquire_ctx *ticket, * to the device. * @res: The resource to make visible to the device. * @intr: Perform waits interruptible if possible. + * @dirtying: Pending GPU operation will dirty the resource * * On succesful return, any backup DMA buffer pointed to by @res->backup will * be reserved and validated. @@ -688,7 +699,8 @@ static int vmw_resource_do_evict(struct ww_acquire_ctx *ticket, * Return: Zero on success, -ERESTARTSYS if interrupted, negative error code * on failure. */ -int vmw_resource_validate(struct vmw_resource *res, bool intr) +int vmw_resource_validate(struct vmw_resource *res, bool intr, + bool dirtying) { int ret; struct vmw_resource *evict_res; @@ -705,7 +717,7 @@ int vmw_resource_validate(struct vmw_resource *res, bool intr) if (res->backup) val_buf.bo = &res->backup->base; do { - ret = vmw_resource_do_validate(res, &val_buf); + ret = vmw_resource_do_validate(res, &val_buf, dirtying); if (likely(ret != -EBUSY)) break; @@ -1005,7 +1017,7 @@ int vmw_resource_pin(struct vmw_resource *res, bool interruptible) /* Do we really need to pin the MOB as well? */ vmw_bo_pin_reserved(vbo, true); } - ret = vmw_resource_validate(res, interruptible); + ret = vmw_resource_validate(res, interruptible, true); if (vbo) ttm_bo_unreserve(&vbo->base); if (ret) @@ -1080,3 +1092,86 @@ void vmw_resource_dirty_update(struct vmw_resource *res, pgoff_t start, res->func->dirty_range_add(res, start << PAGE_SHIFT, end << PAGE_SHIFT); } + +/** + * vmw_resources_clean - Clean resources intersecting a mob range + * @vbo: The mob buffer object + * @start: The mob page offset starting the range + * @end: The mob page offset ending the range + * @num_prefault: Returns how many pages including the first have been + * cleaned and are ok to prefault + */ +int vmw_resources_clean(struct vmw_buffer_object *vbo, pgoff_t start, + pgoff_t end, pgoff_t *num_prefault) +{ + struct rb_node *cur = vbo->res_tree.rb_node; + struct vmw_resource *found = NULL; + unsigned long res_start = start << PAGE_SHIFT; + unsigned long res_end = end << PAGE_SHIFT; + unsigned long last_cleaned = 0; + + /* + * Find the resource with lowest backup_offset that intersects the + * range. + */ + while (cur) { + struct vmw_resource *cur_res = + container_of(cur, struct vmw_resource, mob_node); + + if (cur_res->backup_offset >= res_end) { + cur = cur->rb_left; + } else if (cur_res->backup_offset + cur_res->backup_size <= + res_start) { + cur = cur->rb_right; + } else { + found = cur_res; + cur = cur->rb_left; + /* Continue to look for resources with lower offsets */ + } + } + + /* + * In order of increasing backup_offset, clean dirty resorces + * intersecting the range. + */ + while (found) { + if (found->res_dirty) { + int ret; + + if (!found->func->clean) + return -EINVAL; + + ret = found->func->clean(found); + if (ret) + return ret; + + found->res_dirty = false; + } + last_cleaned = found->backup_offset + found->backup_size; + cur = rb_next(&found->mob_node); + if (!cur) + break; + + found = container_of(cur, struct vmw_resource, mob_node); + if (found->backup_offset >= res_end) + break; + } + + /* + * Set number of pages allowed prefaulting and fence the buffer object + */ + *num_prefault = 1; + if (last_cleaned > res_start) { + struct ttm_buffer_object *bo = &vbo->base; + + *num_prefault = __KERNEL_DIV_ROUND_UP(last_cleaned - res_start, + PAGE_SIZE); + vmw_bo_fence_single(bo, NULL); + if (bo->moving) + dma_fence_put(bo->moving); + bo->moving = dma_fence_get + (dma_resv_get_excl(bo->base.resv)); + } + + return 0; +} diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h index c85144286cfe..3b7438b2d289 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h @@ -77,6 +77,7 @@ struct vmw_user_resource_conv { * @dirty_sync: Upload the dirty mob contents to the resource. * @dirty_add_range: Add a sequential dirty range to the resource * dirty tracker. + * @clean: Clean the resource. */ struct vmw_res_func { enum vmw_res_type res_type; @@ -101,6 +102,7 @@ struct vmw_res_func { int (*dirty_sync)(struct vmw_resource *res); void (*dirty_range_add)(struct vmw_resource *res, size_t start, size_t end); + int (*clean)(struct vmw_resource *res); }; /** diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c index 71349a7bae90..9aaf807ed73c 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c @@ -641,7 +641,8 @@ int vmw_validation_res_validate(struct vmw_validation_context *ctx, bool intr) struct vmw_resource *res = val->res; struct vmw_buffer_object *backup = res->backup; - ret = vmw_resource_validate(res, intr); + ret = vmw_resource_validate(res, intr, val->dirty_set && + val->dirty); if (ret) { if (ret != -ERESTARTSYS) DRM_ERROR("Failed to validate resource.\n"); From patchwork Wed Sep 18 12:59:14 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: =?utf-8?q?Thomas_Hellstr=C3=B6m_=28Intel=29?= X-Patchwork-Id: 11150321 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 10D6C14DB for ; Wed, 18 Sep 2019 12:59:49 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id ED6B7207FC for ; Wed, 18 Sep 2019 12:59:48 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org ED6B7207FC Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=shipmail.org Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=dri-devel-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id F19DA6EF39; Wed, 18 Sep 2019 12:59:44 +0000 (UTC) X-Original-To: dri-devel@lists.freedesktop.org Delivered-To: dri-devel@lists.freedesktop.org Received: from pio-pvt-msa3.bahnhof.se (pio-pvt-msa3.bahnhof.se [79.136.2.42]) by gabe.freedesktop.org (Postfix) with ESMTPS id 30B5C6EF39 for ; Wed, 18 Sep 2019 12:59:38 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by pio-pvt-msa3.bahnhof.se (Postfix) with ESMTP id 62C773F872; Wed, 18 Sep 2019 14:59:31 +0200 (CEST) X-Virus-Scanned: Debian amavisd-new at bahnhof.se X-Spam-Flag: NO X-Spam-Score: -2.099 X-Spam-Level: X-Spam-Status: No, score=-2.099 tagged_above=-999 required=6.31 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no Received: from pio-pvt-msa3.bahnhof.se ([127.0.0.1]) by localhost (pio-pvt-msa3.bahnhof.se [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id vC88il-KpaU4; Wed, 18 Sep 2019 14:59:26 +0200 (CEST) Received: from mail1.shipmail.org (h-205-35.A357.priv.bahnhof.se [155.4.205.35]) (Authenticated sender: mb878879) by pio-pvt-msa3.bahnhof.se (Postfix) with ESMTPA id 0CB283F868; Wed, 18 Sep 2019 14:59:26 +0200 (CEST) Received: from localhost.localdomain.localdomain (h-205-35.A357.priv.bahnhof.se [155.4.205.35]) by mail1.shipmail.org (Postfix) with ESMTPSA id 305CE3605DF; Wed, 18 Sep 2019 14:59:25 +0200 (CEST) From: =?utf-8?q?Thomas_Hellstr=C3=B6m_=28VMware=29?= To: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, linux-mm@kvack.org Subject: [PATCH 7/7] drm/vmwgfx: Add surface dirty-tracking callbacks Date: Wed, 18 Sep 2019 14:59:14 +0200 Message-Id: <20190918125914.38497-8-thomas_os@shipmail.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20190918125914.38497-1-thomas_os@shipmail.org> References: <20190918125914.38497-1-thomas_os@shipmail.org> MIME-Version: 1.0 X-Mailman-Original-DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=shipmail.org; s=mail; t=1568811565; bh=aCZQT2v8Wee6NSIhnf4v4kqWqhCp27Ghfw4lWQiUTQI=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=BShPzOoKeKN+XHXDwmIMwF9OsI/uz986Chjn3FuZ7ND9NOgn860axH4SMB8p++TPr L11+eLDZyp4CJDzRXoe/0zZYeptl7R3ziVuqIMPnTPE4kXJ0JSLyggIgl0kWezwNRs tUYBD9ZLN85hzp57lHgV6mKADnFpY2eXRbI7VI30= X-Mailman-Original-Authentication-Results: pio-pvt-msa3.bahnhof.se; dkim=pass (1024-bit key; unprotected) header.d=shipmail.org header.i=@shipmail.org header.b=BShPzOoK; dkim-atps=neutral X-BeenThere: dri-devel@lists.freedesktop.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Direct Rendering Infrastructure - Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Thomas Hellstrom , Michal Hocko , Rik van Riel , pv-drivers@vmware.com, Minchan Kim , Will Deacon , Matthew Wilcox , Christoph Hellwig , Peter Zijlstra , =?utf-8?b?SsOpcsO0bWUgR2xpc3Nl?= , linux-graphics-maintainer@vmware.com, Souptick Joarder , Huang Ying , Andrew Morton , Deepak Rawat , =?utf-8?q?Christian_K=C3=B6nig?= Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" From: Thomas Hellstrom Add the callbacks necessary to implement emulated coherent memory for surfaces. Add a flag to the gb_surface_create ioctl to indicate that surface memory should be coherent. Also bump the drm minor version to signal the availability of coherent surfaces. Cc: Andrew Morton Cc: Matthew Wilcox Cc: Will Deacon Cc: Peter Zijlstra Cc: Rik van Riel Cc: Minchan Kim Cc: Michal Hocko Cc: Huang Ying Cc: Souptick Joarder Cc: "Jérôme Glisse" Cc: "Christian König" Cc: Christoph Hellwig Signed-off-by: Thomas Hellstrom Reviewed-by: Deepak Rawat --- .../device_include/svga3d_surfacedefs.h | 233 ++++++++++- drivers/gpu/drm/vmwgfx/vmwgfx_drv.h | 4 +- drivers/gpu/drm/vmwgfx/vmwgfx_surface.c | 395 +++++++++++++++++- include/uapi/drm/vmwgfx_drm.h | 4 +- 4 files changed, 629 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/vmwgfx/device_include/svga3d_surfacedefs.h b/drivers/gpu/drm/vmwgfx/device_include/svga3d_surfacedefs.h index f2bfd3d80598..61414f105c67 100644 --- a/drivers/gpu/drm/vmwgfx/device_include/svga3d_surfacedefs.h +++ b/drivers/gpu/drm/vmwgfx/device_include/svga3d_surfacedefs.h @@ -1280,7 +1280,6 @@ svga3dsurface_get_pixel_offset(SVGA3dSurfaceFormat format, return offset; } - static inline u32 svga3dsurface_get_image_offset(SVGA3dSurfaceFormat format, surf_size_struct baseLevelSize, @@ -1375,4 +1374,236 @@ svga3dsurface_is_screen_target_format(SVGA3dSurfaceFormat format) return svga3dsurface_is_dx_screen_target_format(format); } +/** + * struct svga3dsurface_mip - Mimpmap level information + * @bytes: Bytes required in the backing store of this mipmap level. + * @img_stride: Byte stride per image. + * @row_stride: Byte stride per block row. + * @size: The size of the mipmap. + */ +struct svga3dsurface_mip { + size_t bytes; + size_t img_stride; + size_t row_stride; + struct drm_vmw_size size; + +}; + +/** + * struct svga3dsurface_cache - Cached surface information + * @desc: Pointer to the surface descriptor + * @mip: Array of mipmap level information. Valid size is @num_mip_levels. + * @mip_chain_bytes: Bytes required in the backing store for the whole chain + * of mip levels. + * @sheet_bytes: Bytes required in the backing store for a sheet + * representing a single sample. + * @num_mip_levels: Valid size of the @mip array. Number of mipmap levels in + * a chain. + * @num_layers: Number of slices in an array texture or number of faces in + * a cubemap texture. + */ +struct svga3dsurface_cache { + const struct svga3d_surface_desc *desc; + struct svga3dsurface_mip mip[DRM_VMW_MAX_MIP_LEVELS]; + size_t mip_chain_bytes; + size_t sheet_bytes; + u32 num_mip_levels; + u32 num_layers; +}; + +/** + * struct svga3dsurface_loc - Surface location + * @sub_resource: Surface subresource. Defined as layer * num_mip_levels + + * mip_level. + * @x: X coordinate. + * @y: Y coordinate. + * @z: Z coordinate. + */ +struct svga3dsurface_loc { + u32 sub_resource; + u32 x, y, z; +}; + +/** + * svga3dsurface_subres - Compute the subresource from layer and mipmap. + * @cache: Surface layout data. + * @mip_level: The mipmap level. + * @layer: The surface layer (face or array slice). + * + * Return: The subresource. + */ +static inline u32 svga3dsurface_subres(const struct svga3dsurface_cache *cache, + u32 mip_level, u32 layer) +{ + return cache->num_mip_levels * layer + mip_level; +} + +/** + * svga3dsurface_setup_cache - Build a surface cache entry + * @size: The surface base level dimensions. + * @format: The surface format. + * @num_mip_levels: Number of mipmap levels. + * @num_layers: Number of layers. + * @cache: Pointer to a struct svga3dsurface_cach object to be filled in. + * + * Return: Zero on success, -EINVAL on invalid surface layout. + */ +static inline int svga3dsurface_setup_cache(const struct drm_vmw_size *size, + SVGA3dSurfaceFormat format, + u32 num_mip_levels, + u32 num_layers, + u32 num_samples, + struct svga3dsurface_cache *cache) +{ + const struct svga3d_surface_desc *desc; + u32 i; + + memset(cache, 0, sizeof(*cache)); + cache->desc = desc = svga3dsurface_get_desc(format); + cache->num_mip_levels = num_mip_levels; + cache->num_layers = num_layers; + for (i = 0; i < cache->num_mip_levels; i++) { + struct svga3dsurface_mip *mip = &cache->mip[i]; + + mip->size = svga3dsurface_get_mip_size(*size, i); + mip->bytes = svga3dsurface_get_image_buffer_size + (desc, &mip->size, 0); + mip->row_stride = + __KERNEL_DIV_ROUND_UP(mip->size.width, + desc->block_size.width) * + desc->bytes_per_block * num_samples; + if (!mip->row_stride) + goto invalid_dim; + + mip->img_stride = + __KERNEL_DIV_ROUND_UP(mip->size.height, + desc->block_size.height) * + mip->row_stride; + if (!mip->img_stride) + goto invalid_dim; + + cache->mip_chain_bytes += mip->bytes; + } + cache->sheet_bytes = cache->mip_chain_bytes * num_layers; + if (!cache->sheet_bytes) + goto invalid_dim; + + return 0; + +invalid_dim: + VMW_DEBUG_USER("Invalid surface layout for dirty tracking.\n"); + return -EINVAL; +} + +/** + * svga3dsurface_get_loc - Get a surface location from an offset into the + * backing store + * @cache: Surface layout data. + * @loc: Pointer to a struct svga3dsurface_loc to be filled in. + * @offset: Offset into the surface backing store. + */ +static inline void +svga3dsurface_get_loc(const struct svga3dsurface_cache *cache, + struct svga3dsurface_loc *loc, + size_t offset) +{ + const struct svga3dsurface_mip *mip = &cache->mip[0]; + const struct svga3d_surface_desc *desc = cache->desc; + u32 layer; + int i; + + if (offset >= cache->sheet_bytes) + offset %= cache->sheet_bytes; + + layer = offset / cache->mip_chain_bytes; + offset -= layer * cache->mip_chain_bytes; + for (i = 0; i < cache->num_mip_levels; ++i, ++mip) { + if (mip->bytes > offset) + break; + offset -= mip->bytes; + } + + loc->sub_resource = svga3dsurface_subres(cache, i, layer); + loc->z = offset / mip->img_stride; + offset -= loc->z * mip->img_stride; + loc->z *= desc->block_size.depth; + loc->y = offset / mip->row_stride; + offset -= loc->y * mip->row_stride; + loc->y *= desc->block_size.height; + loc->x = offset / desc->bytes_per_block; + loc->x *= desc->block_size.width; +} + +/** + * svga3dsurface_inc_loc - Clamp increment a surface location with one block + * size + * in each dimension. + * @loc: Pointer to a struct svga3dsurface_loc to be incremented. + * + * When computing the size of a range as size = end - start, the range does not + * include the end element. However a location representing the last byte + * of a touched region in the backing store *is* included in the range. + * This function modifies such a location to match the end definition + * given as start + size which is the one used in a SVGA3dBox. + */ +static inline void +svga3dsurface_inc_loc(const struct svga3dsurface_cache *cache, + struct svga3dsurface_loc *loc) +{ + const struct svga3d_surface_desc *desc = cache->desc; + u32 mip = loc->sub_resource % cache->num_mip_levels; + const struct drm_vmw_size *size = &cache->mip[mip].size; + + loc->sub_resource++; + loc->x += desc->block_size.width; + if (loc->x > size->width) + loc->x = size->width; + loc->y += desc->block_size.height; + if (loc->y > size->height) + loc->y = size->height; + loc->z += desc->block_size.depth; + if (loc->z > size->depth) + loc->z = size->depth; +} + +/** + * svga3dsurface_min_loc - The start location in a subresource + * @cache: Surface layout data. + * @sub_resource: The subresource. + * @loc: Pointer to a struct svga3dsurface_loc to be filled in. + */ +static inline void +svga3dsurface_min_loc(const struct svga3dsurface_cache *cache, + u32 sub_resource, + struct svga3dsurface_loc *loc) +{ + loc->sub_resource = sub_resource; + loc->x = loc->y = loc->z = 0; +} + +/** + * svga3dsurface_min_loc - The end location in a subresource + * @cache: Surface layout data. + * @sub_resource: The subresource. + * @loc: Pointer to a struct svga3dsurface_loc to be filled in. + * + * Following the end definition given in svga3dsurface_inc_loc(), + * Compute the end location of a surface subresource. + */ +static inline void +svga3dsurface_max_loc(const struct svga3dsurface_cache *cache, + u32 sub_resource, + struct svga3dsurface_loc *loc) +{ + const struct drm_vmw_size *size; + u32 mip; + + loc->sub_resource = sub_resource + 1; + mip = sub_resource % cache->num_mip_levels; + size = &cache->mip[mip].size; + loc->x = size->width; + loc->y = size->height; + loc->z = size->depth; +} + #endif /* _SVGA3D_SURFACEDEFS_H_ */ diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h index 729a2e93acf1..f5261e1c96d7 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h @@ -56,9 +56,9 @@ #define VMWGFX_DRIVER_NAME "vmwgfx" -#define VMWGFX_DRIVER_DATE "20180704" +#define VMWGFX_DRIVER_DATE "20190328" #define VMWGFX_DRIVER_MAJOR 2 -#define VMWGFX_DRIVER_MINOR 15 +#define VMWGFX_DRIVER_MINOR 16 #define VMWGFX_DRIVER_PATCHLEVEL 0 #define VMWGFX_FIFO_STATIC_SIZE (1024*1024) #define VMWGFX_MAX_RELOCATIONS 2048 diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c index 29d8794f0421..876bada5b35e 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c @@ -68,6 +68,20 @@ struct vmw_surface_offset { uint32_t bo_offset; }; +/** + * vmw_surface_dirty - Surface dirty-tracker + * @cache: Cached layout information of the surface. + * @size: Accounting size for the struct vmw_surface_dirty. + * @num_subres: Number of subresources. + * @boxes: Array of SVGA3dBoxes indicating dirty regions. One per subresource. + */ +struct vmw_surface_dirty { + struct svga3dsurface_cache cache; + size_t size; + u32 num_subres; + SVGA3dBox boxes[0]; +}; + static void vmw_user_surface_free(struct vmw_resource *res); static struct vmw_resource * vmw_user_surface_base_to_res(struct ttm_base_object *base); @@ -96,6 +110,13 @@ vmw_gb_surface_reference_internal(struct drm_device *dev, struct drm_vmw_gb_surface_ref_ext_rep *rep, struct drm_file *file_priv); +static void vmw_surface_dirty_free(struct vmw_resource *res); +static int vmw_surface_dirty_alloc(struct vmw_resource *res); +static int vmw_surface_dirty_sync(struct vmw_resource *res); +static void vmw_surface_dirty_range_add(struct vmw_resource *res, size_t start, + size_t end); +static int vmw_surface_clean(struct vmw_resource *res); + static const struct vmw_user_resource_conv user_surface_conv = { .object_type = VMW_RES_SURFACE, .base_obj_to_res = vmw_user_surface_base_to_res, @@ -133,7 +154,12 @@ static const struct vmw_res_func vmw_gb_surface_func = { .create = vmw_gb_surface_create, .destroy = vmw_gb_surface_destroy, .bind = vmw_gb_surface_bind, - .unbind = vmw_gb_surface_unbind + .unbind = vmw_gb_surface_unbind, + .dirty_alloc = vmw_surface_dirty_alloc, + .dirty_free = vmw_surface_dirty_free, + .dirty_sync = vmw_surface_dirty_sync, + .dirty_range_add = vmw_surface_dirty_range_add, + .clean = vmw_surface_clean, }; /** @@ -641,6 +667,7 @@ static void vmw_user_surface_free(struct vmw_resource *res) struct vmw_private *dev_priv = srf->res.dev_priv; uint32_t size = user_srf->size; + WARN_ON_ONCE(res->dirty); if (user_srf->master) drm_master_put(&user_srf->master); kfree(srf->offsets); @@ -1168,10 +1195,16 @@ static int vmw_gb_surface_bind(struct vmw_resource *res, cmd2->header.id = SVGA_3D_CMD_UPDATE_GB_SURFACE; cmd2->header.size = sizeof(cmd2->body); cmd2->body.sid = res->id; - res->backup_dirty = false; } vmw_fifo_commit(dev_priv, submit_size); + if (res->backup->dirty && res->backup_dirty) { + /* We've just made a full upload. Cear dirty regions. */ + vmw_bo_dirty_clear_res(res); + } + + res->backup_dirty = false; + return 0; } @@ -1636,7 +1669,8 @@ vmw_gb_surface_define_internal(struct drm_device *dev, } } } else if (req->base.drm_surface_flags & - drm_vmw_surface_flag_create_buffer) + (drm_vmw_surface_flag_create_buffer | + drm_vmw_surface_flag_coherent)) ret = vmw_user_bo_alloc(dev_priv, tfile, res->backup_size, req->base.drm_surface_flags & @@ -1650,6 +1684,26 @@ vmw_gb_surface_define_internal(struct drm_device *dev, goto out_unlock; } + if (req->base.drm_surface_flags & drm_vmw_surface_flag_coherent) { + struct vmw_buffer_object *backup = res->backup; + + ttm_bo_reserve(&backup->base, false, false, NULL); + if (!res->func->dirty_alloc) + ret = -EINVAL; + if (!ret) + ret = vmw_bo_dirty_add(backup); + if (!ret) { + res->coherent = true; + ret = res->func->dirty_alloc(res); + } + ttm_bo_unreserve(&backup->base); + if (ret) { + vmw_resource_unreference(&res); + goto out_unlock; + } + + } + tmp = vmw_resource_reference(res); ret = ttm_prime_object_init(tfile, res->backup_size, &user_srf->prime, req->base.drm_surface_flags & @@ -1758,3 +1812,338 @@ vmw_gb_surface_reference_internal(struct drm_device *dev, return ret; } + +/** + * vmw_subres_dirty_add - Add a dirty region to a subresource + * @dirty: The surfaces's dirty tracker. + * @loc_start: The location corresponding to the start of the region. + * @loc_end: The location corresponding to the end of the region. + * + * As we are assuming that @loc_start and @loc_end represent a sequential + * range of backing store memory, if the region spans multiple lines then + * regardless of the x coordinate, the full lines are dirtied. + * Correspondingly if the region spans multiple z slices, then full rather + * than partial z slices are dirtied. + */ +static void vmw_subres_dirty_add(struct vmw_surface_dirty *dirty, + const struct svga3dsurface_loc *loc_start, + const struct svga3dsurface_loc *loc_end) +{ + const struct svga3dsurface_cache *cache = &dirty->cache; + SVGA3dBox *box = &dirty->boxes[loc_start->sub_resource]; + u32 mip = loc_start->sub_resource % cache->num_mip_levels; + const struct drm_vmw_size *size = &cache->mip[mip].size; + u32 box_c2 = box->z + box->d; + + if (WARN_ON(loc_start->sub_resource >= dirty->num_subres)) + return; + + if (box->d == 0 || box->z > loc_start->z) + box->z = loc_start->z; + if (box_c2 < loc_end->z) + box->d = loc_end->z - box->z; + + if (loc_start->z + 1 == loc_end->z) { + box_c2 = box->y + box->h; + if (box->h == 0 || box->y > loc_start->y) + box->y = loc_start->y; + if (box_c2 < loc_end->y) + box->h = loc_end->y - box->y; + + if (loc_start->y + 1 == loc_end->y) { + box_c2 = box->x + box->w; + if (box->w == 0 || box->x > loc_start->x) + box->x = loc_start->x; + if (box_c2 < loc_end->x) + box->w = loc_end->x - box->x; + } else { + box->x = 0; + box->w = size->width; + } + } else { + box->y = 0; + box->h = size->height; + box->x = 0; + box->w = size->width; + } +} + +/** + * vmw_subres_dirty_full - Mark a full subresource as dirty + * @dirty: The surface's dirty tracker. + * @subres: The subresource + */ +static void vmw_subres_dirty_full(struct vmw_surface_dirty *dirty, u32 subres) +{ + const struct svga3dsurface_cache *cache = &dirty->cache; + u32 mip = subres % cache->num_mip_levels; + const struct drm_vmw_size *size = &cache->mip[mip].size; + SVGA3dBox *box = &dirty->boxes[subres]; + + box->x = 0; + box->y = 0; + box->z = 0; + box->w = size->width; + box->h = size->height; + box->d = size->depth; +} + +/* + * vmw_surface_tex_dirty_add_range - The dirty_add_range callback for texture + * surfaces. + */ +static void vmw_surface_tex_dirty_range_add(struct vmw_resource *res, + size_t start, size_t end) +{ + struct vmw_surface_dirty *dirty = + (struct vmw_surface_dirty *) res->dirty; + size_t backup_end = res->backup_offset + res->backup_size; + struct svga3dsurface_loc loc1, loc2; + const struct svga3dsurface_cache *cache; + + start = max_t(size_t, start, res->backup_offset) - res->backup_offset; + end = min(end, backup_end) - res->backup_offset; + cache = &dirty->cache; + svga3dsurface_get_loc(cache, &loc1, start); + svga3dsurface_get_loc(cache, &loc2, end - 1); + svga3dsurface_inc_loc(cache, &loc2); + + if (loc1.sub_resource + 1 == loc2.sub_resource) { + /* Dirty range covers a single sub-resource */ + vmw_subres_dirty_add(dirty, &loc1, &loc2); + } else { + /* Dirty range covers multiple sub-resources */ + struct svga3dsurface_loc loc_min, loc_max; + u32 sub_res = loc1.sub_resource; + + svga3dsurface_max_loc(cache, loc1.sub_resource, &loc_max); + vmw_subres_dirty_add(dirty, &loc1, &loc_max); + svga3dsurface_min_loc(cache, loc2.sub_resource - 1, &loc_min); + vmw_subres_dirty_add(dirty, &loc_min, &loc2); + for (sub_res = loc1.sub_resource + 1; + sub_res < loc2.sub_resource - 1; ++sub_res) + vmw_subres_dirty_full(dirty, sub_res); + } +} + +/* + * vmw_surface_tex_dirty_add_range - The dirty_add_range callback for buffer + * surfaces. + */ +static void vmw_surface_buf_dirty_range_add(struct vmw_resource *res, + size_t start, size_t end) +{ + struct vmw_surface_dirty *dirty = + (struct vmw_surface_dirty *) res->dirty; + const struct svga3dsurface_cache *cache = &dirty->cache; + size_t backup_end = res->backup_offset + cache->mip_chain_bytes; + SVGA3dBox *box = &dirty->boxes[0]; + u32 box_c2; + + box->h = box->d = 1; + start = max_t(size_t, start, res->backup_offset) - res->backup_offset; + end = min(end, backup_end) - res->backup_offset; + box_c2 = box->x + box->w; + if (box->w == 0 || box->x > start) + box->x = start; + if (box_c2 < end) + box->w = end - box->x; +} + +/* + * vmw_surface_tex_dirty_add_range - The dirty_add_range callback for surfaces + */ +static void vmw_surface_dirty_range_add(struct vmw_resource *res, size_t start, + size_t end) +{ + struct vmw_surface *srf = vmw_res_to_srf(res); + + if (WARN_ON(end <= res->backup_offset || + start >= res->backup_offset + res->backup_size)) + return; + + if (srf->format == SVGA3D_BUFFER) + vmw_surface_buf_dirty_range_add(res, start, end); + else + vmw_surface_tex_dirty_range_add(res, start, end); +} + +/* + * vmw_surface_dirty_sync - The surface's dirty_sync callback. + */ +static int vmw_surface_dirty_sync(struct vmw_resource *res) +{ + struct vmw_private *dev_priv = res->dev_priv; + bool has_dx = 0; + u32 i, num_dirty; + struct vmw_surface_dirty *dirty = + (struct vmw_surface_dirty *) res->dirty; + size_t alloc_size; + const struct svga3dsurface_cache *cache = &dirty->cache; + struct { + SVGA3dCmdHeader header; + SVGA3dCmdDXUpdateSubResource body; + } *cmd1; + struct { + SVGA3dCmdHeader header; + SVGA3dCmdUpdateGBImage body; + } *cmd2; + void *cmd; + + num_dirty = 0; + for (i = 0; i < dirty->num_subres; ++i) { + const SVGA3dBox *box = &dirty->boxes[i]; + + if (box->d) + num_dirty++; + } + + if (!num_dirty) + goto out; + + alloc_size = num_dirty * ((has_dx) ? sizeof(*cmd1) : sizeof(*cmd2)); + cmd = VMW_FIFO_RESERVE(dev_priv, alloc_size); + if (!cmd) + return -ENOMEM; + + cmd1 = cmd; + cmd2 = cmd; + + for (i = 0; i < dirty->num_subres; ++i) { + const SVGA3dBox *box = &dirty->boxes[i]; + + if (!box->d) + continue; + + /* + * DX_UPDATE_SUBRESOURCE is aware of array surfaces. + * UPDATE_GB_IMAGE is not. + */ + if (has_dx) { + cmd1->header.id = SVGA_3D_CMD_DX_UPDATE_SUBRESOURCE; + cmd1->header.size = sizeof(cmd1->body); + cmd1->body.sid = res->id; + cmd1->body.subResource = i; + cmd1->body.box = *box; + cmd1++; + } else { + cmd2->header.id = SVGA_3D_CMD_UPDATE_GB_IMAGE; + cmd2->header.size = sizeof(cmd2->body); + cmd2->body.image.sid = res->id; + cmd2->body.image.face = i / cache->num_mip_levels; + cmd2->body.image.mipmap = i - + (cache->num_mip_levels * cmd2->body.image.face); + cmd2->body.box = *box; + cmd2++; + } + + } + vmw_fifo_commit(dev_priv, alloc_size); + out: + memset(&dirty->boxes[0], 0, sizeof(dirty->boxes[0]) * + dirty->num_subres); + + return 0; +} + +/* + * vmw_surface_dirty_alloc - The surface's dirty_alloc callback. + */ +static int vmw_surface_dirty_alloc(struct vmw_resource *res) +{ + struct vmw_surface *srf = vmw_res_to_srf(res); + struct vmw_surface_dirty *dirty; + u32 num_layers = 1; + u32 num_mip; + u32 num_subres; + u32 num_samples; + size_t dirty_size, acc_size; + static struct ttm_operation_ctx ctx = { + .interruptible = false, + .no_wait_gpu = false + }; + int ret; + + if (srf->array_size) + num_layers = srf->array_size; + else if (srf->flags & SVGA3D_SURFACE_CUBEMAP) + num_layers *= SVGA3D_MAX_SURFACE_FACES; + + num_mip = srf->mip_levels[0]; + if (!num_mip) + num_mip = 1; + + num_subres = num_layers * num_mip; + dirty_size = sizeof(*dirty) + num_subres * sizeof(dirty->boxes[0]); + acc_size = ttm_round_pot(dirty_size); + ret = ttm_mem_global_alloc(vmw_mem_glob(res->dev_priv), + acc_size, &ctx); + if (ret) { + VMW_DEBUG_USER("Out of graphics memory for surface " + "dirty tracker.\n"); + return ret; + } + + dirty = kvzalloc(dirty_size, GFP_KERNEL); + if (!dirty) { + ret = -ENOMEM; + goto out_no_dirty; + } + + num_samples = max_t(u32, 1, srf->multisample_count); + ret = svga3dsurface_setup_cache(&srf->base_size, srf->format, num_mip, + num_layers, num_samples, &dirty->cache); + if (ret) + goto out_no_cache; + + dirty->num_subres = num_subres; + dirty->size = acc_size; + res->dirty = (struct vmw_resource_dirty *) dirty; + + return 0; + +out_no_cache: + kvfree(dirty); +out_no_dirty: + ttm_mem_global_free(vmw_mem_glob(res->dev_priv), acc_size); + return ret; +} + +/* + * vmw_surface_dirty_free - The surface's dirty_free callback + */ +static void vmw_surface_dirty_free(struct vmw_resource *res) +{ + struct vmw_surface_dirty *dirty = + (struct vmw_surface_dirty *) res->dirty; + size_t acc_size = dirty->size; + + kvfree(dirty); + ttm_mem_global_free(vmw_mem_glob(res->dev_priv), acc_size); + res->dirty = NULL; +} + +/* + * vmw_surface_clean - The surface's clean callback + */ +static int vmw_surface_clean(struct vmw_resource *res) +{ + struct vmw_private *dev_priv = res->dev_priv; + size_t alloc_size; + struct { + SVGA3dCmdHeader header; + SVGA3dCmdReadbackGBSurface body; + } *cmd; + + alloc_size = sizeof(*cmd); + cmd = VMW_FIFO_RESERVE(dev_priv, alloc_size); + if (!cmd) + return -ENOMEM; + + cmd->header.id = SVGA_3D_CMD_READBACK_GB_SURFACE; + cmd->header.size = sizeof(cmd->body); + cmd->body.sid = res->id; + vmw_fifo_commit(dev_priv, alloc_size); + + return 0; +} diff --git a/include/uapi/drm/vmwgfx_drm.h b/include/uapi/drm/vmwgfx_drm.h index 399f58317cff..02cab33f2f25 100644 --- a/include/uapi/drm/vmwgfx_drm.h +++ b/include/uapi/drm/vmwgfx_drm.h @@ -891,11 +891,13 @@ struct drm_vmw_shader_arg { * surface. * @drm_vmw_surface_flag_create_buffer: Create a backup buffer if none is * given. + * @drm_vmw_surface_flag_coherent: Back surface with coherent memory. */ enum drm_vmw_surface_flags { drm_vmw_surface_flag_shareable = (1 << 0), drm_vmw_surface_flag_scanout = (1 << 1), - drm_vmw_surface_flag_create_buffer = (1 << 2) + drm_vmw_surface_flag_create_buffer = (1 << 2), + drm_vmw_surface_flag_coherent = (1 << 3), }; /**