From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-7.0 required=3.0 tests=DKIM_SIGNED,DKIM_VALID, DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY, SPF_PASS,T_MIXED_ES,URIBL_BLOCKED autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id AEACAC65BAE for ; Thu, 13 Dec 2018 10:17:02 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 4982B21104 for ; Thu, 13 Dec 2018 10:17:02 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.b="TO8h9/Zy" DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 4982B21104 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=linux-kernel-owner@vger.kernel.org Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728217AbeLMKRA (ORCPT ); Thu, 13 Dec 2018 05:17:00 -0500 Received: from mail-lj1-f193.google.com ([209.85.208.193]:34638 "EHLO mail-lj1-f193.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727991AbeLMKRA (ORCPT ); Thu, 13 Dec 2018 05:17:00 -0500 Received: by mail-lj1-f193.google.com with SMTP id u6-v6so1294896ljd.1 for ; Thu, 13 Dec 2018 02:16:56 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=subject:from:to:cc:references:message-id:date:user-agent :mime-version:in-reply-to:content-transfer-encoding:content-language; bh=12Z5Ye4CSFcoUec9aRHmkE4BwAzVYrZVzk1V/J8URjw=; b=TO8h9/ZypOSETP9tZ2YcoCKBVH2Ph6TNGdQiSULejdyM70N7NLR2eiJYD1xA9DojpP /MfemPJLWFMwhxrre0XVdDQLjZq9dVj6eSzIojg/9mSAB5tgxIUzfri4Yge5TswVo3w8 boRMlJsZp1QSo/0+vhzBJD4DBTO99Q9oQY4LaOJfx34zj5fvqftk6GRPwhra5DKtD8Kl lxE/yC/SW7+UYQXc+YOj4tLbd6pIm+rs1njCvpbnVTw/ua2D7TTLgqKkxGrtTGFDWuza 36WPF32mgp1EhbR7iP4JDr26GxEVsTBi3Gskpl9YqV1qPpi+FQM0bjKw+dxe1m2WzEYz JMzA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:subject:from:to:cc:references:message-id:date :user-agent:mime-version:in-reply-to:content-transfer-encoding :content-language; bh=12Z5Ye4CSFcoUec9aRHmkE4BwAzVYrZVzk1V/J8URjw=; b=sPlurPWqtcORKwWAxmjzGCvMZ+/ny99U5xxWt9jIt2+p+JjnOej6KUrSP2DJ3Nvphc TSg6B6c5Dv5VPm+pSnWozxHbRS5xXPR+3gBx16Rz5GLwRzaX6lx0+Xx7+ny/tEDB8hc2 nIDYC1vvUal+ovWTKrqygzuEiWy6VXTyIsrg4G1x4CFSg2SeGSNpyUUWGuhC3Htjlcmv SH4cnFAac2BNo8wafOgIVwm4iqbnelSkSPk1UInBszU+Z3eRMEMp1Gmk8k2XA26IbWu/ ZShy7z5ksIYLfwN+Ak0WwczBymuW8IL9hzUgUFVcZ/h7JvMK8PiHsUzeg69KMvh0YdLt 5v6A== X-Gm-Message-State: AA+aEWZNC2lrIqHPiIHzV46C2L4C+w/tYPQTwC6NuGLjlHNGfZTQK9JT ryzFvIUwdhevQpKC9cu7150= X-Google-Smtp-Source: AFSGD/VsgM/p8I+Vvk4I1MLmmh034J0g9ugaMWtX4G/hpRC9Ww2vGEV15PLWyjsOqYvMJZDOO6OtDA== X-Received: by 2002:a2e:80d3:: with SMTP id r19-v6mr14385625ljg.151.1544696214868; Thu, 13 Dec 2018 02:16:54 -0800 (PST) Received: from [10.17.182.20] (ll-22.209.223.85.sovam.net.ua. [85.223.209.22]) by smtp.gmail.com with ESMTPSA id g17sm260702lfg.78.2018.12.13.02.16.53 (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Thu, 13 Dec 2018 02:16:54 -0800 (PST) Subject: Re: [Xen-devel][PATCH v2 2/3] drm/xen-front: Use Xen common shared buffer implementation From: Oleksandr Andrushchenko To: Daniel Vetter Cc: xen-devel@lists.xenproject.org, linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, alsa-devel@alsa-project.org, jgross@suse.com, boris.ostrovsky@oracle.com, Oleksandr Andrushchenko References: <20181130074205.20496-1-andr2000@gmail.com> <20181130074205.20496-2-andr2000@gmail.com> <3c69d924-11e1-7037-bfa5-276d22b245fe@gmail.com> Message-ID: <27491aa5-3548-4c39-13a2-132c738fae14@gmail.com> Date: Thu, 13 Dec 2018 12:16:52 +0200 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.2.1 MIME-Version: 1.0 In-Reply-To: <3c69d924-11e1-7037-bfa5-276d22b245fe@gmail.com> Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 8bit Content-Language: en-US Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org bump On 12/5/18 10:20 AM, Oleksandr Andrushchenko wrote: > Hello, Daniel! > > Could you please ack/nack the patch, so either we can merge the > > series or I can address your comments if any > > Thank you, > > Oleksandr > > On 11/30/18 9:42 AM, Oleksandr Andrushchenko wrote: >> From: Oleksandr Andrushchenko >> >> Use page directory based shared buffer implementation >> now available as common code for Xen frontend drivers. >> >> Remove flushing of shared buffer on page flip as this >> workaround needs a proper fix. >> >> Signed-off-by: Oleksandr Andrushchenko >> >> --- >>   drivers/gpu/drm/xen/Kconfig               |   1 + >>   drivers/gpu/drm/xen/Makefile              |   1 - >>   drivers/gpu/drm/xen/xen_drm_front.c       |  65 ++-- >>   drivers/gpu/drm/xen/xen_drm_front_gem.c   |   1 - >>   drivers/gpu/drm/xen/xen_drm_front_shbuf.c | 414 ---------------------- >>   drivers/gpu/drm/xen/xen_drm_front_shbuf.h |  64 ---- >>   6 files changed, 26 insertions(+), 520 deletions(-) >>   delete mode 100644 drivers/gpu/drm/xen/xen_drm_front_shbuf.c >>   delete mode 100644 drivers/gpu/drm/xen/xen_drm_front_shbuf.h >> >> diff --git a/drivers/gpu/drm/xen/Kconfig b/drivers/gpu/drm/xen/Kconfig >> index 4cca160782ab..f969d486855d 100644 >> --- a/drivers/gpu/drm/xen/Kconfig >> +++ b/drivers/gpu/drm/xen/Kconfig >> @@ -12,6 +12,7 @@ config DRM_XEN_FRONTEND >>       select DRM_KMS_HELPER >>       select VIDEOMODE_HELPERS >>       select XEN_XENBUS_FRONTEND >> +    select XEN_FRONT_PGDIR_SHBUF >>       help >>         Choose this option if you want to enable a para-virtualized >>         frontend DRM/KMS driver for Xen guest OSes. >> diff --git a/drivers/gpu/drm/xen/Makefile b/drivers/gpu/drm/xen/Makefile >> index 712afff5ffc3..825905f67faa 100644 >> --- a/drivers/gpu/drm/xen/Makefile >> +++ b/drivers/gpu/drm/xen/Makefile >> @@ -4,7 +4,6 @@ drm_xen_front-objs := xen_drm_front.o \ >>                 xen_drm_front_kms.o \ >>                 xen_drm_front_conn.o \ >>                 xen_drm_front_evtchnl.o \ >> -              xen_drm_front_shbuf.o \ >>                 xen_drm_front_cfg.o \ >>                 xen_drm_front_gem.o >>   diff --git a/drivers/gpu/drm/xen/xen_drm_front.c >> b/drivers/gpu/drm/xen/xen_drm_front.c >> index 6b6d5ab82ec3..4d3d36fc3a5d 100644 >> --- a/drivers/gpu/drm/xen/xen_drm_front.c >> +++ b/drivers/gpu/drm/xen/xen_drm_front.c >> @@ -19,6 +19,7 @@ >>   #include >>   #include >>   +#include >>   #include >>     #include "xen_drm_front.h" >> @@ -26,28 +27,20 @@ >>   #include "xen_drm_front_evtchnl.h" >>   #include "xen_drm_front_gem.h" >>   #include "xen_drm_front_kms.h" >> -#include "xen_drm_front_shbuf.h" >>     struct xen_drm_front_dbuf { >>       struct list_head list; >>       u64 dbuf_cookie; >>       u64 fb_cookie; >> -    struct xen_drm_front_shbuf *shbuf; >> + >> +    struct xen_front_pgdir_shbuf shbuf; >>   }; >>   -static int dbuf_add_to_list(struct xen_drm_front_info *front_info, >> -                struct xen_drm_front_shbuf *shbuf, u64 dbuf_cookie) >> +static void dbuf_add_to_list(struct xen_drm_front_info *front_info, >> +                 struct xen_drm_front_dbuf *dbuf, u64 dbuf_cookie) >>   { >> -    struct xen_drm_front_dbuf *dbuf; >> - >> -    dbuf = kzalloc(sizeof(*dbuf), GFP_KERNEL); >> -    if (!dbuf) >> -        return -ENOMEM; >> - >>       dbuf->dbuf_cookie = dbuf_cookie; >> -    dbuf->shbuf = shbuf; >>       list_add(&dbuf->list, &front_info->dbuf_list); >> -    return 0; >>   } >>     static struct xen_drm_front_dbuf *dbuf_get(struct list_head >> *dbuf_list, >> @@ -62,15 +55,6 @@ static struct xen_drm_front_dbuf *dbuf_get(struct >> list_head *dbuf_list, >>       return NULL; >>   } >>   -static void dbuf_flush_fb(struct list_head *dbuf_list, u64 fb_cookie) >> -{ >> -    struct xen_drm_front_dbuf *buf, *q; >> - >> -    list_for_each_entry_safe(buf, q, dbuf_list, list) >> -        if (buf->fb_cookie == fb_cookie) >> -            xen_drm_front_shbuf_flush(buf->shbuf); >> -} >> - >>   static void dbuf_free(struct list_head *dbuf_list, u64 dbuf_cookie) >>   { >>       struct xen_drm_front_dbuf *buf, *q; >> @@ -78,8 +62,8 @@ static void dbuf_free(struct list_head *dbuf_list, >> u64 dbuf_cookie) >>       list_for_each_entry_safe(buf, q, dbuf_list, list) >>           if (buf->dbuf_cookie == dbuf_cookie) { >>               list_del(&buf->list); >> -            xen_drm_front_shbuf_unmap(buf->shbuf); >> -            xen_drm_front_shbuf_free(buf->shbuf); >> +            xen_front_pgdir_shbuf_unmap(&buf->shbuf); >> +            xen_front_pgdir_shbuf_free(&buf->shbuf); >>               kfree(buf); >>               break; >>           } >> @@ -91,8 +75,8 @@ static void dbuf_free_all(struct list_head *dbuf_list) >>         list_for_each_entry_safe(buf, q, dbuf_list, list) { >>           list_del(&buf->list); >> -        xen_drm_front_shbuf_unmap(buf->shbuf); >> -        xen_drm_front_shbuf_free(buf->shbuf); >> +        xen_front_pgdir_shbuf_unmap(&buf->shbuf); >> +        xen_front_pgdir_shbuf_free(&buf->shbuf); >>           kfree(buf); >>       } >>   } >> @@ -171,9 +155,9 @@ int xen_drm_front_dbuf_create(struct >> xen_drm_front_info *front_info, >>                     u32 bpp, u64 size, struct page **pages) >>   { >>       struct xen_drm_front_evtchnl *evtchnl; >> -    struct xen_drm_front_shbuf *shbuf; >> +    struct xen_drm_front_dbuf *dbuf; >>       struct xendispl_req *req; >> -    struct xen_drm_front_shbuf_cfg buf_cfg; >> +    struct xen_front_pgdir_shbuf_cfg buf_cfg; >>       unsigned long flags; >>       int ret; >>   @@ -181,28 +165,29 @@ int xen_drm_front_dbuf_create(struct >> xen_drm_front_info *front_info, >>       if (unlikely(!evtchnl)) >>           return -EIO; >>   +    dbuf = kzalloc(sizeof(*dbuf), GFP_KERNEL); >> +    if (!dbuf) >> +        return -ENOMEM; >> + >> +    dbuf_add_to_list(front_info, dbuf, dbuf_cookie); >> + >>       memset(&buf_cfg, 0, sizeof(buf_cfg)); >>       buf_cfg.xb_dev = front_info->xb_dev; >> +    buf_cfg.num_pages = DIV_ROUND_UP(size, PAGE_SIZE); >>       buf_cfg.pages = pages; >> -    buf_cfg.size = size; >> +    buf_cfg.pgdir = &dbuf->shbuf; >>       buf_cfg.be_alloc = front_info->cfg.be_alloc; >>   -    shbuf = xen_drm_front_shbuf_alloc(&buf_cfg); >> -    if (IS_ERR(shbuf)) >> -        return PTR_ERR(shbuf); >> - >> -    ret = dbuf_add_to_list(front_info, shbuf, dbuf_cookie); >> -    if (ret < 0) { >> -        xen_drm_front_shbuf_free(shbuf); >> -        return ret; >> -    } >> +    ret = xen_front_pgdir_shbuf_alloc(&buf_cfg); >> +    if (ret < 0) >> +        goto fail_shbuf_alloc; >>         mutex_lock(&evtchnl->u.req.req_io_lock); >>         spin_lock_irqsave(&front_info->io_lock, flags); >>       req = be_prepare_req(evtchnl, XENDISPL_OP_DBUF_CREATE); >>       req->op.dbuf_create.gref_directory = >> -            xen_drm_front_shbuf_get_dir_start(shbuf); >> + xen_front_pgdir_shbuf_get_dir_start(&dbuf->shbuf); >>       req->op.dbuf_create.buffer_sz = size; >>       req->op.dbuf_create.dbuf_cookie = dbuf_cookie; >>       req->op.dbuf_create.width = width; >> @@ -221,7 +206,7 @@ int xen_drm_front_dbuf_create(struct >> xen_drm_front_info *front_info, >>       if (ret < 0) >>           goto fail; >>   -    ret = xen_drm_front_shbuf_map(shbuf); >> +    ret = xen_front_pgdir_shbuf_map(&dbuf->shbuf); >>       if (ret < 0) >>           goto fail; >>   @@ -230,6 +215,7 @@ int xen_drm_front_dbuf_create(struct >> xen_drm_front_info *front_info, >>     fail: >>       mutex_unlock(&evtchnl->u.req.req_io_lock); >> +fail_shbuf_alloc: >>       dbuf_free(&front_info->dbuf_list, dbuf_cookie); >>       return ret; >>   } >> @@ -358,7 +344,6 @@ int xen_drm_front_page_flip(struct >> xen_drm_front_info *front_info, >>       if (unlikely(conn_idx >= front_info->num_evt_pairs)) >>           return -EINVAL; >>   -    dbuf_flush_fb(&front_info->dbuf_list, fb_cookie); >>       evtchnl = &front_info->evt_pairs[conn_idx].req; >>         mutex_lock(&evtchnl->u.req.req_io_lock); >> diff --git a/drivers/gpu/drm/xen/xen_drm_front_gem.c >> b/drivers/gpu/drm/xen/xen_drm_front_gem.c >> index c85bfe7571cb..53be4c7dc0a4 100644 >> --- a/drivers/gpu/drm/xen/xen_drm_front_gem.c >> +++ b/drivers/gpu/drm/xen/xen_drm_front_gem.c >> @@ -22,7 +22,6 @@ >>   #include >>     #include "xen_drm_front.h" >> -#include "xen_drm_front_shbuf.h" >>     struct xen_gem_object { >>       struct drm_gem_object base; >> diff --git a/drivers/gpu/drm/xen/xen_drm_front_shbuf.c >> b/drivers/gpu/drm/xen/xen_drm_front_shbuf.c >> deleted file mode 100644 >> index d333b67cc1a0..000000000000 >> --- a/drivers/gpu/drm/xen/xen_drm_front_shbuf.c >> +++ /dev/null >> @@ -1,414 +0,0 @@ >> -// SPDX-License-Identifier: GPL-2.0 OR MIT >> - >> -/* >> - *  Xen para-virtual DRM device >> - * >> - * Copyright (C) 2016-2018 EPAM Systems Inc. >> - * >> - * Author: Oleksandr Andrushchenko >> - */ >> - >> -#include >> - >> -#if defined(CONFIG_X86) >> -#include >> -#endif >> -#include >> -#include >> - >> -#include >> -#include >> -#include >> -#include >> -#include >> -#include >> - >> -#include "xen_drm_front.h" >> -#include "xen_drm_front_shbuf.h" >> - >> -struct xen_drm_front_shbuf_ops { >> -    /* >> -     * Calculate number of grefs required to handle this buffer, >> -     * e.g. if grefs are required for page directory only or the buffer >> -     * pages as well. >> -     */ >> -    void (*calc_num_grefs)(struct xen_drm_front_shbuf *buf); >> -    /* Fill page directory according to para-virtual display >> protocol. */ >> -    void (*fill_page_dir)(struct xen_drm_front_shbuf *buf); >> -    /* Claim grant references for the pages of the buffer. */ >> -    int (*grant_refs_for_buffer)(struct xen_drm_front_shbuf *buf, >> -                     grant_ref_t *priv_gref_head, int gref_idx); >> -    /* Map grant references of the buffer. */ >> -    int (*map)(struct xen_drm_front_shbuf *buf); >> -    /* Unmap grant references of the buffer. */ >> -    int (*unmap)(struct xen_drm_front_shbuf *buf); >> -}; >> - >> -grant_ref_t xen_drm_front_shbuf_get_dir_start(struct >> xen_drm_front_shbuf *buf) >> -{ >> -    if (!buf->grefs) >> -        return GRANT_INVALID_REF; >> - >> -    return buf->grefs[0]; >> -} >> - >> -int xen_drm_front_shbuf_map(struct xen_drm_front_shbuf *buf) >> -{ >> -    if (buf->ops->map) >> -        return buf->ops->map(buf); >> - >> -    /* no need to map own grant references */ >> -    return 0; >> -} >> - >> -int xen_drm_front_shbuf_unmap(struct xen_drm_front_shbuf *buf) >> -{ >> -    if (buf->ops->unmap) >> -        return buf->ops->unmap(buf); >> - >> -    /* no need to unmap own grant references */ >> -    return 0; >> -} >> - >> -void xen_drm_front_shbuf_flush(struct xen_drm_front_shbuf *buf) >> -{ >> -#if defined(CONFIG_X86) >> -    drm_clflush_pages(buf->pages, buf->num_pages); >> -#endif >> -} >> - >> -void xen_drm_front_shbuf_free(struct xen_drm_front_shbuf *buf) >> -{ >> -    if (buf->grefs) { >> -        int i; >> - >> -        for (i = 0; i < buf->num_grefs; i++) >> -            if (buf->grefs[i] != GRANT_INVALID_REF) >> -                gnttab_end_foreign_access(buf->grefs[i], >> -                              0, 0UL); >> -    } >> -    kfree(buf->grefs); >> -    kfree(buf->directory); >> -    kfree(buf); >> -} >> - >> -/* >> - * number of grefs a page can hold with respect to the >> - * struct xendispl_page_directory header >> - */ >> -#define XEN_DRM_NUM_GREFS_PER_PAGE ((PAGE_SIZE - \ >> -        offsetof(struct xendispl_page_directory, gref)) / \ >> -        sizeof(grant_ref_t)) >> - >> -static int get_num_pages_dir(struct xen_drm_front_shbuf *buf) >> -{ >> -    /* number of pages the page directory consumes itself */ >> -    return DIV_ROUND_UP(buf->num_pages, XEN_DRM_NUM_GREFS_PER_PAGE); >> -} >> - >> -static void backend_calc_num_grefs(struct xen_drm_front_shbuf *buf) >> -{ >> -    /* only for pages the page directory consumes itself */ >> -    buf->num_grefs = get_num_pages_dir(buf); >> -} >> - >> -static void guest_calc_num_grefs(struct xen_drm_front_shbuf *buf) >> -{ >> -    /* >> -     * number of pages the page directory consumes itself >> -     * plus grefs for the buffer pages >> -     */ >> -    buf->num_grefs = get_num_pages_dir(buf) + buf->num_pages; >> -} >> - >> -#define xen_page_to_vaddr(page) \ >> -        ((uintptr_t)pfn_to_kaddr(page_to_xen_pfn(page))) >> - >> -static int backend_unmap(struct xen_drm_front_shbuf *buf) >> -{ >> -    struct gnttab_unmap_grant_ref *unmap_ops; >> -    int i, ret; >> - >> -    if (!buf->pages || !buf->backend_map_handles || !buf->grefs) >> -        return 0; >> - >> -    unmap_ops = kcalloc(buf->num_pages, sizeof(*unmap_ops), >> -                GFP_KERNEL); >> -    if (!unmap_ops) { >> -        DRM_ERROR("Failed to get memory while unmapping\n"); >> -        return -ENOMEM; >> -    } >> - >> -    for (i = 0; i < buf->num_pages; i++) { >> -        phys_addr_t addr; >> - >> -        addr = xen_page_to_vaddr(buf->pages[i]); >> -        gnttab_set_unmap_op(&unmap_ops[i], addr, GNTMAP_host_map, >> -                    buf->backend_map_handles[i]); >> -    } >> - >> -    ret = gnttab_unmap_refs(unmap_ops, NULL, buf->pages, >> -                buf->num_pages); >> - >> -    for (i = 0; i < buf->num_pages; i++) { >> -        if (unlikely(unmap_ops[i].status != GNTST_okay)) >> -            DRM_ERROR("Failed to unmap page %d: %d\n", >> -                  i, unmap_ops[i].status); >> -    } >> - >> -    if (ret) >> -        DRM_ERROR("Failed to unmap grant references, ret %d", ret); >> - >> -    kfree(unmap_ops); >> -    kfree(buf->backend_map_handles); >> -    buf->backend_map_handles = NULL; >> -    return ret; >> -} >> - >> -static int backend_map(struct xen_drm_front_shbuf *buf) >> -{ >> -    struct gnttab_map_grant_ref *map_ops = NULL; >> -    unsigned char *ptr; >> -    int ret, cur_gref, cur_dir_page, cur_page, grefs_left; >> - >> -    map_ops = kcalloc(buf->num_pages, sizeof(*map_ops), GFP_KERNEL); >> -    if (!map_ops) >> -        return -ENOMEM; >> - >> -    buf->backend_map_handles = kcalloc(buf->num_pages, >> -                       sizeof(*buf->backend_map_handles), >> -                       GFP_KERNEL); >> -    if (!buf->backend_map_handles) { >> -        kfree(map_ops); >> -        return -ENOMEM; >> -    } >> - >> -    /* >> -     * read page directory to get grefs from the backend: for external >> -     * buffer we only allocate buf->grefs for the page directory, >> -     * so buf->num_grefs has number of pages in the page directory >> itself >> -     */ >> -    ptr = buf->directory; >> -    grefs_left = buf->num_pages; >> -    cur_page = 0; >> -    for (cur_dir_page = 0; cur_dir_page < buf->num_grefs; >> cur_dir_page++) { >> -        struct xendispl_page_directory *page_dir = >> -                (struct xendispl_page_directory *)ptr; >> -        int to_copy = XEN_DRM_NUM_GREFS_PER_PAGE; >> - >> -        if (to_copy > grefs_left) >> -            to_copy = grefs_left; >> - >> -        for (cur_gref = 0; cur_gref < to_copy; cur_gref++) { >> -            phys_addr_t addr; >> - >> -            addr = xen_page_to_vaddr(buf->pages[cur_page]); >> -            gnttab_set_map_op(&map_ops[cur_page], addr, >> -                      GNTMAP_host_map, >> -                      page_dir->gref[cur_gref], >> -                      buf->xb_dev->otherend_id); >> -            cur_page++; >> -        } >> - >> -        grefs_left -= to_copy; >> -        ptr += PAGE_SIZE; >> -    } >> -    ret = gnttab_map_refs(map_ops, NULL, buf->pages, buf->num_pages); >> - >> -    /* save handles even if error, so we can unmap */ >> -    for (cur_page = 0; cur_page < buf->num_pages; cur_page++) { >> -        buf->backend_map_handles[cur_page] = map_ops[cur_page].handle; >> -        if (unlikely(map_ops[cur_page].status != GNTST_okay)) >> -            DRM_ERROR("Failed to map page %d: %d\n", >> -                  cur_page, map_ops[cur_page].status); >> -    } >> - >> -    if (ret) { >> -        DRM_ERROR("Failed to map grant references, ret %d", ret); >> -        backend_unmap(buf); >> -    } >> - >> -    kfree(map_ops); >> -    return ret; >> -} >> - >> -static void backend_fill_page_dir(struct xen_drm_front_shbuf *buf) >> -{ >> -    struct xendispl_page_directory *page_dir; >> -    unsigned char *ptr; >> -    int i, num_pages_dir; >> - >> -    ptr = buf->directory; >> -    num_pages_dir = get_num_pages_dir(buf); >> - >> -    /* fill only grefs for the page directory itself */ >> -    for (i = 0; i < num_pages_dir - 1; i++) { >> -        page_dir = (struct xendispl_page_directory *)ptr; >> - >> -        page_dir->gref_dir_next_page = buf->grefs[i + 1]; >> -        ptr += PAGE_SIZE; >> -    } >> -    /* last page must say there is no more pages */ >> -    page_dir = (struct xendispl_page_directory *)ptr; >> -    page_dir->gref_dir_next_page = GRANT_INVALID_REF; >> -} >> - >> -static void guest_fill_page_dir(struct xen_drm_front_shbuf *buf) >> -{ >> -    unsigned char *ptr; >> -    int cur_gref, grefs_left, to_copy, i, num_pages_dir; >> - >> -    ptr = buf->directory; >> -    num_pages_dir = get_num_pages_dir(buf); >> - >> -    /* >> -     * while copying, skip grefs at start, they are for pages >> -     * granted for the page directory itself >> -     */ >> -    cur_gref = num_pages_dir; >> -    grefs_left = buf->num_pages; >> -    for (i = 0; i < num_pages_dir; i++) { >> -        struct xendispl_page_directory *page_dir = >> -                (struct xendispl_page_directory *)ptr; >> - >> -        if (grefs_left <= XEN_DRM_NUM_GREFS_PER_PAGE) { >> -            to_copy = grefs_left; >> -            page_dir->gref_dir_next_page = GRANT_INVALID_REF; >> -        } else { >> -            to_copy = XEN_DRM_NUM_GREFS_PER_PAGE; >> -            page_dir->gref_dir_next_page = buf->grefs[i + 1]; >> -        } >> -        memcpy(&page_dir->gref, &buf->grefs[cur_gref], >> -               to_copy * sizeof(grant_ref_t)); >> -        ptr += PAGE_SIZE; >> -        grefs_left -= to_copy; >> -        cur_gref += to_copy; >> -    } >> -} >> - >> -static int guest_grant_refs_for_buffer(struct xen_drm_front_shbuf *buf, >> -                       grant_ref_t *priv_gref_head, >> -                       int gref_idx) >> -{ >> -    int i, cur_ref, otherend_id; >> - >> -    otherend_id = buf->xb_dev->otherend_id; >> -    for (i = 0; i < buf->num_pages; i++) { >> -        cur_ref = gnttab_claim_grant_reference(priv_gref_head); >> -        if (cur_ref < 0) >> -            return cur_ref; >> - >> -        gnttab_grant_foreign_access_ref(cur_ref, otherend_id, >> -                        xen_page_to_gfn(buf->pages[i]), >> -                        0); >> -        buf->grefs[gref_idx++] = cur_ref; >> -    } >> -    return 0; >> -} >> - >> -static int grant_references(struct xen_drm_front_shbuf *buf) >> -{ >> -    grant_ref_t priv_gref_head; >> -    int ret, i, j, cur_ref; >> -    int otherend_id, num_pages_dir; >> - >> -    ret = gnttab_alloc_grant_references(buf->num_grefs, >> &priv_gref_head); >> -    if (ret < 0) { >> -        DRM_ERROR("Cannot allocate grant references\n"); >> -        return ret; >> -    } >> - >> -    otherend_id = buf->xb_dev->otherend_id; >> -    j = 0; >> -    num_pages_dir = get_num_pages_dir(buf); >> -    for (i = 0; i < num_pages_dir; i++) { >> -        unsigned long frame; >> - >> -        cur_ref = gnttab_claim_grant_reference(&priv_gref_head); >> -        if (cur_ref < 0) >> -            return cur_ref; >> - >> -        frame = xen_page_to_gfn(virt_to_page(buf->directory + >> -                    PAGE_SIZE * i)); >> -        gnttab_grant_foreign_access_ref(cur_ref, otherend_id, frame, >> 0); >> -        buf->grefs[j++] = cur_ref; >> -    } >> - >> -    if (buf->ops->grant_refs_for_buffer) { >> -        ret = buf->ops->grant_refs_for_buffer(buf, &priv_gref_head, j); >> -        if (ret) >> -            return ret; >> -    } >> - >> -    gnttab_free_grant_references(priv_gref_head); >> -    return 0; >> -} >> - >> -static int alloc_storage(struct xen_drm_front_shbuf *buf) >> -{ >> -    buf->grefs = kcalloc(buf->num_grefs, sizeof(*buf->grefs), >> GFP_KERNEL); >> -    if (!buf->grefs) >> -        return -ENOMEM; >> - >> -    buf->directory = kcalloc(get_num_pages_dir(buf), PAGE_SIZE, >> GFP_KERNEL); >> -    if (!buf->directory) >> -        return -ENOMEM; >> - >> -    return 0; >> -} >> - >> -/* >> - * For be allocated buffers we don't need grant_refs_for_buffer as >> those >> - * grant references are allocated at backend side >> - */ >> -static const struct xen_drm_front_shbuf_ops backend_ops = { >> -    .calc_num_grefs = backend_calc_num_grefs, >> -    .fill_page_dir = backend_fill_page_dir, >> -    .map = backend_map, >> -    .unmap = backend_unmap >> -}; >> - >> -/* For locally granted references we do not need to map/unmap the >> references */ >> -static const struct xen_drm_front_shbuf_ops local_ops = { >> -    .calc_num_grefs = guest_calc_num_grefs, >> -    .fill_page_dir = guest_fill_page_dir, >> -    .grant_refs_for_buffer = guest_grant_refs_for_buffer, >> -}; >> - >> -struct xen_drm_front_shbuf * >> -xen_drm_front_shbuf_alloc(struct xen_drm_front_shbuf_cfg *cfg) >> -{ >> -    struct xen_drm_front_shbuf *buf; >> -    int ret; >> - >> -    buf = kzalloc(sizeof(*buf), GFP_KERNEL); >> -    if (!buf) >> -        return ERR_PTR(-ENOMEM); >> - >> -    if (cfg->be_alloc) >> -        buf->ops = &backend_ops; >> -    else >> -        buf->ops = &local_ops; >> - >> -    buf->xb_dev = cfg->xb_dev; >> -    buf->num_pages = DIV_ROUND_UP(cfg->size, PAGE_SIZE); >> -    buf->pages = cfg->pages; >> - >> -    buf->ops->calc_num_grefs(buf); >> - >> -    ret = alloc_storage(buf); >> -    if (ret) >> -        goto fail; >> - >> -    ret = grant_references(buf); >> -    if (ret) >> -        goto fail; >> - >> -    buf->ops->fill_page_dir(buf); >> - >> -    return buf; >> - >> -fail: >> -    xen_drm_front_shbuf_free(buf); >> -    return ERR_PTR(ret); >> -} >> diff --git a/drivers/gpu/drm/xen/xen_drm_front_shbuf.h >> b/drivers/gpu/drm/xen/xen_drm_front_shbuf.h >> deleted file mode 100644 >> index 7545c692539e..000000000000 >> --- a/drivers/gpu/drm/xen/xen_drm_front_shbuf.h >> +++ /dev/null >> @@ -1,64 +0,0 @@ >> -/* SPDX-License-Identifier: GPL-2.0 OR MIT */ >> - >> -/* >> - *  Xen para-virtual DRM device >> - * >> - * Copyright (C) 2016-2018 EPAM Systems Inc. >> - * >> - * Author: Oleksandr Andrushchenko >> - */ >> - >> -#ifndef __XEN_DRM_FRONT_SHBUF_H_ >> -#define __XEN_DRM_FRONT_SHBUF_H_ >> - >> -#include >> -#include >> - >> -#include >> - >> -struct xen_drm_front_shbuf { >> -    /* >> -     * number of references granted for the backend use: >> -     *  - for allocated/imported dma-buf's this holds number of grant >> -     *    references for the page directory and pages of the buffer >> -     *  - for the buffer provided by the backend this holds number of >> -     *    grant references for the page directory as grant >> references for >> -     *    the buffer will be provided by the backend >> -     */ >> -    int num_grefs; >> -    grant_ref_t *grefs; >> -    unsigned char *directory; >> - >> -    int num_pages; >> -    struct page **pages; >> - >> -    struct xenbus_device *xb_dev; >> - >> -    /* these are the ops used internally depending on be_alloc mode */ >> -    const struct xen_drm_front_shbuf_ops *ops; >> - >> -    /* Xen map handles for the buffer allocated by the backend */ >> -    grant_handle_t *backend_map_handles; >> -}; >> - >> -struct xen_drm_front_shbuf_cfg { >> -    struct xenbus_device *xb_dev; >> -    size_t size; >> -    struct page **pages; >> -    bool be_alloc; >> -}; >> - >> -struct xen_drm_front_shbuf * >> -xen_drm_front_shbuf_alloc(struct xen_drm_front_shbuf_cfg *cfg); >> - >> -grant_ref_t xen_drm_front_shbuf_get_dir_start(struct >> xen_drm_front_shbuf *buf); >> - >> -int xen_drm_front_shbuf_map(struct xen_drm_front_shbuf *buf); >> - >> -int xen_drm_front_shbuf_unmap(struct xen_drm_front_shbuf *buf); >> - >> -void xen_drm_front_shbuf_flush(struct xen_drm_front_shbuf *buf); >> - >> -void xen_drm_front_shbuf_free(struct xen_drm_front_shbuf *buf); >> - >> -#endif /* __XEN_DRM_FRONT_SHBUF_H_ */