]> www.infradead.org Git - users/willy/pagecache.git/commitdiff
Drivers: hv: Cosmetic changes for hv.c and balloon.c
authorAditya Nagesh <adityanagesh@linux.microsoft.com>
Fri, 31 May 2024 10:48:41 +0000 (03:48 -0700)
committerWei Liu <wei.liu@kernel.org>
Thu, 6 Jun 2024 06:03:29 +0000 (06:03 +0000)
Fix issues reported by checkpatch.pl script in hv.c and
balloon.c
 - Remove unnecessary parentheses
 - Remove extra newlines
 - Remove extra spaces
 - Add spaces between comparison operators
 - Remove comparison with NULL in if statements

No functional changes intended

Signed-off-by: Aditya Nagesh <adityanagesh@linux.microsoft.com>
Reviewed-by: Saurabh Sengar <ssengar@linux.microsoft.com>
Reviewed-by: Michael Kelley <mhklinux@outlook.com>
Link: https://lore.kernel.org/r/1717152521-6439-1-git-send-email-adityanagesh@linux.microsoft.com
Signed-off-by: Wei Liu <wei.liu@kernel.org>
Message-ID: <1717152521-6439-1-git-send-email-adityanagesh@linux.microsoft.com>

drivers/hv/hv.c
drivers/hv/hv_balloon.c

index a8ad728354cb097621db497cd48251861a0f1016..e0d676c74f1478eb7478313ee79dd6bf8aee3225 100644 (file)
@@ -45,8 +45,8 @@ int hv_init(void)
  * This involves a hypercall.
  */
 int hv_post_message(union hv_connection_id connection_id,
-                 enum hv_message_type message_type,
-                 void *payload, size_t payload_size)
+                       enum hv_message_type message_type,
+                       void *payload, size_t payload_size)
 {
        struct hv_input_post_message *aligned_msg;
        unsigned long flags;
@@ -86,7 +86,7 @@ int hv_post_message(union hv_connection_id connection_id,
                        status = HV_STATUS_INVALID_PARAMETER;
        } else {
                status = hv_do_hypercall(HVCALL_POST_MESSAGE,
-                               aligned_msg, NULL);
+                                        aligned_msg, NULL);
        }
 
        local_irq_restore(flags);
@@ -111,7 +111,7 @@ int hv_synic_alloc(void)
 
        hv_context.hv_numa_map = kcalloc(nr_node_ids, sizeof(struct cpumask),
                                         GFP_KERNEL);
-       if (hv_context.hv_numa_map == NULL) {
+       if (!hv_context.hv_numa_map) {
                pr_err("Unable to allocate NUMA map\n");
                goto err;
        }
@@ -120,11 +120,11 @@ int hv_synic_alloc(void)
                hv_cpu = per_cpu_ptr(hv_context.cpu_context, cpu);
 
                tasklet_init(&hv_cpu->msg_dpc,
-                            vmbus_on_msg_dpc, (unsigned long) hv_cpu);
+                            vmbus_on_msg_dpc, (unsigned long)hv_cpu);
 
                if (ms_hyperv.paravisor_present && hv_isolation_type_tdx()) {
                        hv_cpu->post_msg_page = (void *)get_zeroed_page(GFP_ATOMIC);
-                       if (hv_cpu->post_msg_page == NULL) {
+                       if (!hv_cpu->post_msg_page) {
                                pr_err("Unable to allocate post msg page\n");
                                goto err;
                        }
@@ -147,14 +147,14 @@ int hv_synic_alloc(void)
                if (!ms_hyperv.paravisor_present && !hv_root_partition) {
                        hv_cpu->synic_message_page =
                                (void *)get_zeroed_page(GFP_ATOMIC);
-                       if (hv_cpu->synic_message_page == NULL) {
+                       if (!hv_cpu->synic_message_page) {
                                pr_err("Unable to allocate SYNIC message page\n");
                                goto err;
                        }
 
                        hv_cpu->synic_event_page =
                                (void *)get_zeroed_page(GFP_ATOMIC);
-                       if (hv_cpu->synic_event_page == NULL) {
+                       if (!hv_cpu->synic_event_page) {
                                pr_err("Unable to allocate SYNIC event page\n");
 
                                free_page((unsigned long)hv_cpu->synic_message_page);
@@ -203,14 +203,13 @@ err:
        return ret;
 }
 
-
 void hv_synic_free(void)
 {
        int cpu, ret;
 
        for_each_present_cpu(cpu) {
-               struct hv_per_cpu_context *hv_cpu
-                       per_cpu_ptr(hv_context.cpu_context, cpu);
+               struct hv_per_cpu_context *hv_cpu =
+                       per_cpu_ptr(hv_context.cpu_context, cpu);
 
                /* It's better to leak the page if the encryption fails. */
                if (ms_hyperv.paravisor_present && hv_isolation_type_tdx()) {
@@ -262,8 +261,8 @@ void hv_synic_free(void)
  */
 void hv_synic_enable_regs(unsigned int cpu)
 {
-       struct hv_per_cpu_context *hv_cpu
-               per_cpu_ptr(hv_context.cpu_context, cpu);
+       struct hv_per_cpu_context *hv_cpu =
+               per_cpu_ptr(hv_context.cpu_context, cpu);
        union hv_synic_simp simp;
        union hv_synic_siefp siefp;
        union hv_synic_sint shared_sint;
@@ -277,8 +276,8 @@ void hv_synic_enable_regs(unsigned int cpu)
                /* Mask out vTOM bit. ioremap_cache() maps decrypted */
                u64 base = (simp.base_simp_gpa << HV_HYP_PAGE_SHIFT) &
                                ~ms_hyperv.shared_gpa_boundary;
-               hv_cpu->synic_message_page
-                       (void *)ioremap_cache(base, HV_HYP_PAGE_SIZE);
+               hv_cpu->synic_message_page =
+                       (void *)ioremap_cache(base, HV_HYP_PAGE_SIZE);
                if (!hv_cpu->synic_message_page)
                        pr_err("Fail to map synic message page.\n");
        } else {
@@ -296,8 +295,8 @@ void hv_synic_enable_regs(unsigned int cpu)
                /* Mask out vTOM bit. ioremap_cache() maps decrypted */
                u64 base = (siefp.base_siefp_gpa << HV_HYP_PAGE_SHIFT) &
                                ~ms_hyperv.shared_gpa_boundary;
-               hv_cpu->synic_event_page
-                       (void *)ioremap_cache(base, HV_HYP_PAGE_SIZE);
+               hv_cpu->synic_event_page =
+                       (void *)ioremap_cache(base, HV_HYP_PAGE_SIZE);
                if (!hv_cpu->synic_event_page)
                        pr_err("Fail to map synic event page.\n");
        } else {
@@ -348,8 +347,8 @@ int hv_synic_init(unsigned int cpu)
  */
 void hv_synic_disable_regs(unsigned int cpu)
 {
-       struct hv_per_cpu_context *hv_cpu
-               per_cpu_ptr(hv_context.cpu_context, cpu);
+       struct hv_per_cpu_context *hv_cpu =
+               per_cpu_ptr(hv_context.cpu_context, cpu);
        union hv_synic_sint shared_sint;
        union hv_synic_simp simp;
        union hv_synic_siefp siefp;
index 4370ad31b5b381fd27c93489328adaecea2352ab..0e7427c2baf588af1e9aa3b43931490361c20e68 100644 (file)
@@ -42,8 +42,6 @@
  * Begin protocol definitions.
  */
 
-
-
 /*
  * Protocol versions. The low word is the minor version, the high word the major
  * version.
@@ -72,8 +70,6 @@ enum {
        DYNMEM_PROTOCOL_VERSION_CURRENT = DYNMEM_PROTOCOL_VERSION_WIN10
 };
 
-
-
 /*
  * Message Types
  */
@@ -102,7 +98,6 @@ enum dm_message_type {
        DM_VERSION_1_MAX                = 12
 };
 
-
 /*
  * Structures defining the dynamic memory management
  * protocol.
@@ -116,7 +111,6 @@ union dm_version {
        __u32 version;
 } __packed;
 
-
 union dm_caps {
        struct {
                __u64 balloon:1;
@@ -149,8 +143,6 @@ union dm_mem_page_range {
        __u64  page_range;
 } __packed;
 
-
-
 /*
  * The header for all dynamic memory messages:
  *
@@ -175,7 +167,6 @@ struct dm_message {
        __u8 data[]; /* enclosed message */
 } __packed;
 
-
 /*
  * Specific message types supporting the dynamic memory protocol.
  */
@@ -272,7 +263,6 @@ struct dm_status {
        __u32 io_diff;
 } __packed;
 
-
 /*
  * Message to ask the guest to allocate memory - balloon up message.
  * This message is sent from the host to the guest. The guest may not be
@@ -287,14 +277,13 @@ struct dm_balloon {
        __u32 reservedz;
 } __packed;
 
-
 /*
  * Balloon response message; this message is sent from the guest
  * to the host in response to the balloon message.
  *
  * reservedz: Reserved; must be set to zero.
  * more_pages: If FALSE, this is the last message of the transaction.
- * if TRUE there will atleast one more message from the guest.
+ * if TRUE there will be at least one more message from the guest.
  *
  * range_count: The number of ranges in the range array.
  *
@@ -315,7 +304,7 @@ struct dm_balloon_response {
  * to the guest to give guest more memory.
  *
  * more_pages: If FALSE, this is the last message of the transaction.
- * if TRUE there will atleast one more message from the guest.
+ * if TRUE there will be at least one more message from the guest.
  *
  * reservedz: Reserved; must be set to zero.
  *
@@ -343,7 +332,6 @@ struct dm_unballoon_response {
        struct dm_header hdr;
 } __packed;
 
-
 /*
  * Hot add request message. Message sent from the host to the guest.
  *
@@ -391,7 +379,6 @@ enum dm_info_type {
        MAX_INFO_TYPE
 };
 
-
 /*
  * Header for the information message.
  */
@@ -481,10 +468,10 @@ static unsigned long last_post_time;
 
 static int hv_hypercall_multi_failure;
 
-module_param(hot_add, bool, (S_IRUGO | S_IWUSR));
+module_param(hot_add, bool, 0644);
 MODULE_PARM_DESC(hot_add, "If set attempt memory hot_add");
 
-module_param(pressure_report_delay, uint, (S_IRUGO | S_IWUSR));
+module_param(pressure_report_delay, uint, 0644);
 MODULE_PARM_DESC(pressure_report_delay, "Delay in secs in reporting pressure");
 static atomic_t trans_id = ATOMIC_INIT(0);
 
@@ -503,7 +490,6 @@ enum hv_dm_state {
        DM_INIT_ERROR
 };
 
-
 static __u8 recv_buffer[HV_HYP_PAGE_SIZE];
 static __u8 balloon_up_send_buffer[HV_HYP_PAGE_SIZE];
 
@@ -599,12 +585,12 @@ static inline bool has_pfn_is_backed(struct hv_hotadd_state *has,
        struct hv_hotadd_gap *gap;
 
        /* The page is not backed. */
-       if ((pfn < has->covered_start_pfn) || (pfn >= has->covered_end_pfn))
+       if (pfn < has->covered_start_pfn || pfn >= has->covered_end_pfn)
                return false;
 
        /* Check for gaps. */
        list_for_each_entry(gap, &has->gap_list, list) {
-               if ((pfn >= gap->start_pfn) && (pfn < gap->end_pfn))
+               if (pfn >= gap->start_pfn && pfn < gap->end_pfn)
                        return false;
        }
 
@@ -784,8 +770,8 @@ static void hv_online_page(struct page *pg, unsigned int order)
        guard(spinlock_irqsave)(&dm_device.ha_lock);
        list_for_each_entry(has, &dm_device.ha_region_list, list) {
                /* The page belongs to a different HAS. */
-               if ((pfn < has->start_pfn) ||
-                               (pfn + (1UL << order) > has->end_pfn))
+               if (pfn < has->start_pfn ||
+                   (pfn + (1UL << order) > has->end_pfn))
                        continue;
 
                hv_bring_pgs_online(has, pfn, 1UL << order);
@@ -846,7 +832,7 @@ static int pfn_covered(unsigned long start_pfn, unsigned long pfn_cnt)
 }
 
 static unsigned long handle_pg_range(unsigned long pg_start,
-                                       unsigned long pg_count)
+                                    unsigned long pg_count)
 {
        unsigned long start_pfn = pg_start;
        unsigned long pfn_cnt = pg_count;
@@ -857,7 +843,7 @@ static unsigned long handle_pg_range(unsigned long pg_start,
        unsigned long res = 0, flags;
 
        pr_debug("Hot adding %lu pages starting at pfn 0x%lx.\n", pg_count,
-               pg_start);
+                pg_start);
 
        spin_lock_irqsave(&dm_device.ha_lock, flags);
        list_for_each_entry(has, &dm_device.ha_region_list, list) {
@@ -893,10 +879,9 @@ static unsigned long handle_pg_range(unsigned long pg_start,
                        if (start_pfn > has->start_pfn &&
                            online_section_nr(pfn_to_section_nr(start_pfn)))
                                hv_bring_pgs_online(has, start_pfn, pgs_ol);
-
                }
 
-               if ((has->ha_end_pfn < has->end_pfn) && (pfn_cnt > 0)) {
+               if (has->ha_end_pfn < has->end_pfn && pfn_cnt > 0) {
                        /*
                         * We have some residual hot add range
                         * that needs to be hot added; hot add
@@ -999,7 +984,7 @@ static void hot_add_req(struct work_struct *dummy)
        rg_start = dm->ha_wrk.ha_region_range.finfo.start_page;
        rg_sz = dm->ha_wrk.ha_region_range.finfo.page_cnt;
 
-       if ((rg_start == 0) && (!dm->host_specified_ha_region)) {
+       if (rg_start == 0 && !dm->host_specified_ha_region) {
                /*
                 * The host has not specified the hot-add region.
                 * Based on the hot-add page range being specified,
@@ -1013,7 +998,7 @@ static void hot_add_req(struct work_struct *dummy)
 
        if (do_hot_add)
                resp.page_count = process_hot_add(pg_start, pfn_cnt,
-                                               rg_start, rg_sz);
+                                                 rg_start, rg_sz);
 
        dm->num_pages_added += resp.page_count;
 #endif
@@ -1191,11 +1176,10 @@ static void post_status(struct hv_dynmem_device *dm)
                                sizeof(struct dm_status),
                                (unsigned long)NULL,
                                VM_PKT_DATA_INBAND, 0);
-
 }
 
 static void free_balloon_pages(struct hv_dynmem_device *dm,
-                        union dm_mem_page_range *range_array)
+                              union dm_mem_page_range *range_array)
 {
        int num_pages = range_array->finfo.page_cnt;
        __u64 start_frame = range_array->finfo.start_page;
@@ -1211,8 +1195,6 @@ static void free_balloon_pages(struct hv_dynmem_device *dm,
        }
 }
 
-
-
 static unsigned int alloc_balloon_pages(struct hv_dynmem_device *dm,
                                        unsigned int num_pages,
                                        struct dm_balloon_response *bl_resp,
@@ -1258,7 +1240,6 @@ static unsigned int alloc_balloon_pages(struct hv_dynmem_device *dm,
                        page_to_pfn(pg);
                bl_resp->range_array[i].finfo.page_cnt = alloc_unit;
                bl_resp->hdr.size += sizeof(union dm_mem_page_range);
-
        }
 
        return i * alloc_unit;
@@ -1312,7 +1293,7 @@ static void balloon_up(struct work_struct *dummy)
 
                if (num_ballooned == 0 || num_ballooned == num_pages) {
                        pr_debug("Ballooned %u out of %u requested pages.\n",
-                               num_pages, dm_device.balloon_wrk.num_pages);
+                                num_pages, dm_device.balloon_wrk.num_pages);
 
                        bl_resp->more_pages = 0;
                        done = true;
@@ -1346,16 +1327,15 @@ static void balloon_up(struct work_struct *dummy)
 
                        for (i = 0; i < bl_resp->range_count; i++)
                                free_balloon_pages(&dm_device,
-                                                &bl_resp->range_array[i]);
+                                                  &bl_resp->range_array[i]);
 
                        done = true;
                }
        }
-
 }
 
 static void balloon_down(struct hv_dynmem_device *dm,
-                       struct dm_unballoon_request *req)
+                        struct dm_unballoon_request *req)
 {
        union dm_mem_page_range *range_array = req->range_array;
        int range_count = req->range_count;
@@ -1369,7 +1349,7 @@ static void balloon_down(struct hv_dynmem_device *dm,
        }
 
        pr_debug("Freed %u ballooned pages.\n",
-               prev_pages_ballooned - dm->num_pages_ballooned);
+                prev_pages_ballooned - dm->num_pages_ballooned);
 
        if (req->more_pages == 1)
                return;
@@ -1394,8 +1374,7 @@ static int dm_thread_func(void *dm_dev)
        struct hv_dynmem_device *dm = dm_dev;
 
        while (!kthread_should_stop()) {
-               wait_for_completion_interruptible_timeout(
-                                               &dm_device.config_event, 1*HZ);
+               wait_for_completion_interruptible_timeout(&dm_device.config_event, 1 * HZ);
                /*
                 * The host expects us to post information on the memory
                 * pressure every second.
@@ -1419,9 +1398,8 @@ static int dm_thread_func(void *dm_dev)
        return 0;
 }
 
-
 static void version_resp(struct hv_dynmem_device *dm,
-                       struct dm_version_response *vresp)
+                        struct dm_version_response *vresp)
 {
        struct dm_version_request version_req;
        int ret;
@@ -1482,7 +1460,7 @@ version_error:
 }
 
 static void cap_resp(struct hv_dynmem_device *dm,
-                       struct dm_capabilities_resp_msg *cap_resp)
+                    struct dm_capabilities_resp_msg *cap_resp)
 {
        if (!cap_resp->is_accepted) {
                pr_err("Capabilities not accepted by host\n");
@@ -1515,7 +1493,7 @@ static void balloon_onchannelcallback(void *context)
                switch (dm_hdr->type) {
                case DM_VERSION_RESPONSE:
                        version_resp(dm,
-                                (struct dm_version_response *)dm_msg);
+                                    (struct dm_version_response *)dm_msg);
                        break;
 
                case DM_CAPABILITIES_RESPONSE:
@@ -1545,7 +1523,7 @@ static void balloon_onchannelcallback(void *context)
 
                        dm->state = DM_BALLOON_DOWN;
                        balloon_down(dm,
-                                (struct dm_unballoon_request *)recv_buffer);
+                                    (struct dm_unballoon_request *)recv_buffer);
                        break;
 
                case DM_MEM_HOT_ADD_REQUEST:
@@ -1583,17 +1561,15 @@ static void balloon_onchannelcallback(void *context)
 
                default:
                        pr_warn_ratelimited("Unhandled message: type: %d\n", dm_hdr->type);
-
                }
        }
-
 }
 
 #define HV_LARGE_REPORTING_ORDER       9
 #define HV_LARGE_REPORTING_LEN (HV_HYP_PAGE_SIZE << \
                HV_LARGE_REPORTING_ORDER)
 static int hv_free_page_report(struct page_reporting_dev_info *pr_dev_info,
-                   struct scatterlist *sgl, unsigned int nents)
+                              struct scatterlist *sgl, unsigned int nents)
 {
        unsigned long flags;
        struct hv_memory_hint *hint;
@@ -1628,7 +1604,7 @@ static int hv_free_page_report(struct page_reporting_dev_info *pr_dev_info,
                 */
 
                /* page reporting for pages 2MB or higher */
-               if (order >= HV_LARGE_REPORTING_ORDER ) {
+               if (order >= HV_LARGE_REPORTING_ORDER) {
                        range->page.largepage = 1;
                        range->page_size = HV_GPA_PAGE_RANGE_PAGE_SIZE_2MB;
                        range->base_large_pfn = page_to_hvpfn(
@@ -1642,23 +1618,21 @@ static int hv_free_page_report(struct page_reporting_dev_info *pr_dev_info,
                        range->page.additional_pages =
                                (sg->length / HV_HYP_PAGE_SIZE) - 1;
                }
-
        }
 
        status = hv_do_rep_hypercall(HV_EXT_CALL_MEMORY_HEAT_HINT, nents, 0,
                                     hint, NULL);
        local_irq_restore(flags);
        if (!hv_result_success(status)) {
-
                pr_err("Cold memory discard hypercall failed with status %llx\n",
-                               status);
+                      status);
                if (hv_hypercall_multi_failure > 0)
                        hv_hypercall_multi_failure++;
 
                if (hv_result(status) == HV_STATUS_INVALID_PARAMETER) {
                        pr_err("Underlying Hyper-V does not support order less than 9. Hypercall failed\n");
                        pr_err("Defaulting to page_reporting_order %d\n",
-                                       pageblock_order);
+                              pageblock_order);
                        page_reporting_order = pageblock_order;
                        hv_hypercall_multi_failure++;
                        return -EINVAL;
@@ -1692,7 +1666,7 @@ static void enable_page_reporting(void)
                pr_err("Failed to enable cold memory discard: %d\n", ret);
        } else {
                pr_info("Cold memory discard hint enabled with order %d\n",
-                               page_reporting_order);
+                       page_reporting_order);
        }
 }
 
@@ -1775,7 +1749,7 @@ static int balloon_connect_vsp(struct hv_device *dev)
        if (ret)
                goto out;
 
-       t = wait_for_completion_timeout(&dm_device.host_event, 5*HZ);
+       t = wait_for_completion_timeout(&dm_device.host_event, 5 * HZ);
        if (t == 0) {
                ret = -ETIMEDOUT;
                goto out;
@@ -1833,7 +1807,7 @@ static int balloon_connect_vsp(struct hv_device *dev)
        if (ret)
                goto out;
 
-       t = wait_for_completion_timeout(&dm_device.host_event, 5*HZ);
+       t = wait_for_completion_timeout(&dm_device.host_event, 5 * HZ);
        if (t == 0) {
                ret = -ETIMEDOUT;
                goto out;
@@ -1874,8 +1848,8 @@ static int hv_balloon_debug_show(struct seq_file *f, void *offset)
        char *sname;
 
        seq_printf(f, "%-22s: %u.%u\n", "host_version",
-                               DYNMEM_MAJOR_VERSION(dm->version),
-                               DYNMEM_MINOR_VERSION(dm->version));
+                       DYNMEM_MAJOR_VERSION(dm->version),
+                       DYNMEM_MINOR_VERSION(dm->version));
 
        seq_printf(f, "%-22s:", "capabilities");
        if (ballooning_enabled())
@@ -1924,10 +1898,10 @@ static int hv_balloon_debug_show(struct seq_file *f, void *offset)
        seq_printf(f, "%-22s: %u\n", "pages_ballooned", dm->num_pages_ballooned);
 
        seq_printf(f, "%-22s: %lu\n", "total_pages_committed",
-                               get_pages_committed(dm));
+                  get_pages_committed(dm));
 
        seq_printf(f, "%-22s: %llu\n", "max_dynamic_page_count",
-                               dm->max_dynamic_page_count);
+                  dm->max_dynamic_page_count);
 
        return 0;
 }
@@ -1937,7 +1911,7 @@ DEFINE_SHOW_ATTRIBUTE(hv_balloon_debug);
 static void  hv_balloon_debugfs_init(struct hv_dynmem_device *b)
 {
        debugfs_create_file("hv-balloon", 0444, NULL, b,
-                       &hv_balloon_debug_fops);
+                           &hv_balloon_debug_fops);
 }
 
 static void  hv_balloon_debugfs_exit(struct hv_dynmem_device *b)
@@ -2095,7 +2069,6 @@ static int balloon_suspend(struct hv_device *hv_dev)
        tasklet_enable(&hv_dev->channel->callback_event);
 
        return 0;
-
 }
 
 static int balloon_resume(struct hv_device *dev)
@@ -2154,7 +2127,6 @@ static  struct hv_driver balloon_drv = {
 
 static int __init init_balloon_drv(void)
 {
-
        return vmbus_driver_register(&balloon_drv);
 }