nvdimm.lists.linux.dev archive mirror
 help / color / mirror / Atom feed
From: Ben Widawsky <ben.widawsky@intel.com>
To: Vishal Verma <vishal.l.verma@intel.com>
Cc: linux-cxl@vger.kernel.org, linux-nvdimm@lists.01.org
Subject: Re: [ndctl PATCH v2 03/13] libcxl: add support for command query and submission
Date: Mon, 22 Feb 2021 13:55:39 -0800	[thread overview]
Message-ID: <20210222215539.s3n6yj5lgmdohypt@intel.com> (raw)
In-Reply-To: <20210219020331.725687-4-vishal.l.verma@intel.com>

On 21-02-18 19:03:21, Vishal Verma wrote:
> Add a set of APIs around 'cxl_cmd' for querying the kernel for supported
> commands, allocating and validating command structures against the
> supported set, and submitting the commands.
> 
> 'Query Commands' and 'Send Command' are implemented as IOCTLs in the
> kernel. 'Query Commands' returns information about each supported
> command, such as flags governing its use, or input and output payload
> sizes. This information is used to validate command support, as well as
> set up input and output buffers for command submission.
> 
> Cc: Ben Widawsky <ben.widawsky@intel.com>
> Cc: Dan Williams <dan.j.williams@intel.com>
> Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
> ---
>  cxl/lib/private.h  |  33 ++++
>  cxl/lib/libcxl.c   | 393 +++++++++++++++++++++++++++++++++++++++++++++
>  cxl/libcxl.h       |  11 ++
>  cxl/lib/libcxl.sym |  13 ++
>  4 files changed, 450 insertions(+)
> 
> diff --git a/cxl/lib/private.h b/cxl/lib/private.h
> index fc88fa1..87ca17e 100644
> --- a/cxl/lib/private.h
> +++ b/cxl/lib/private.h
> @@ -4,6 +4,9 @@
>  #define _LIBCXL_PRIVATE_H_
>  
>  #include <libkmod.h>
> +#include <cxl/cxl_mem.h>
> +#include <ccan/endian/endian.h>
> +#include <ccan/short_types/short_types.h>
>  
>  #define CXL_EXPORT __attribute__ ((visibility("default")))
>  
> @@ -21,6 +24,36 @@ struct cxl_memdev {
>  	struct kmod_module *module;
>  };
>  
> +enum cxl_cmd_query_status {
> +	CXL_CMD_QUERY_NOT_RUN = 0,
> +	CXL_CMD_QUERY_OK,
> +	CXL_CMD_QUERY_UNSUPPORTED,
> +};
> +
> +/**
> + * struct cxl_cmd - CXL memdev command
> + * @memdev: the memory device to which the command is being sent
> + * @query_cmd: structure for the Linux 'Query commands' ioctl
> + * @send_cmd: structure for the Linux 'Send command' ioctl
> + * @input_payload: buffer for input payload managed by libcxl
> + * @output_payload: buffer for output payload managed by libcxl
> + * @refcount: reference for passing command buffer around
> + * @query_status: status from query_commands
> + * @query_idx: index of 'this' command in the query_commands array
> + * @status: command return status from the device
> + */
> +struct cxl_cmd {
> +	struct cxl_memdev *memdev;
> +	struct cxl_mem_query_commands *query_cmd;
> +	struct cxl_send_command *send_cmd;
> +	void *input_payload;
> +	void *output_payload;
> +	int refcount;
> +	int query_status;
> +	int query_idx;
> +	int status;
> +};
> +
>  static inline int check_kmod(struct kmod_ctx *kmod_ctx)
>  {
>  	return kmod_ctx ? 0 : -ENXIO;
> diff --git a/cxl/lib/libcxl.c b/cxl/lib/libcxl.c
> index d34e7d0..f1155a1 100644
> --- a/cxl/lib/libcxl.c
> +++ b/cxl/lib/libcxl.c
> @@ -9,6 +9,7 @@
>  #include <unistd.h>
>  #include <sys/stat.h>
>  #include <sys/types.h>
> +#include <sys/ioctl.h>
>  #include <sys/sysmacros.h>
>  #include <uuid/uuid.h>
>  #include <ccan/list/list.h>
> @@ -17,6 +18,7 @@
>  #include <util/log.h>
>  #include <util/sysfs.h>
>  #include <util/bitmap.h>
> +#include <cxl/cxl_mem.h>
>  #include <cxl/libcxl.h>
>  #include "private.h"
>  
> @@ -343,3 +345,394 @@ CXL_EXPORT const char *cxl_memdev_get_firmware_verison(struct cxl_memdev *memdev
>  {
>  	return memdev->firmware_version;
>  }
> +
> +CXL_EXPORT void cxl_cmd_unref(struct cxl_cmd *cmd)
> +{
> +	if (!cmd)
> +		return;
> +	if (--cmd->refcount == 0) {
> +		free(cmd->query_cmd);
> +		free(cmd->send_cmd);
> +		free(cmd->input_payload);
> +		free(cmd->output_payload);
> +		free(cmd);
> +	}
> +}
> +
> +CXL_EXPORT void cxl_cmd_ref(struct cxl_cmd *cmd)
> +{
> +	cmd->refcount++;
> +}
> +
> +static int cxl_cmd_alloc_query(struct cxl_cmd *cmd, int num_cmds)
> +{
> +	size_t size;
> +
> +	if (!cmd)
> +		return -EINVAL;
> +
> +	if (cmd->query_cmd != NULL)
> +		free(cmd->query_cmd);
> +
> +	size = sizeof(struct cxl_mem_query_commands) +
> +			(num_cmds * sizeof(struct cxl_command_info));
> +	cmd->query_cmd = calloc(1, size);
> +	if (!cmd->query_cmd)
> +		return -ENOMEM;
> +
> +	cmd->query_cmd->n_commands = num_cmds;
> +
> +	return 0;
> +}
> +
> +static struct cxl_cmd *cxl_cmd_new(struct cxl_memdev *memdev)
> +{
> +	struct cxl_cmd *cmd;
> +	size_t size;
> +
> +	size = sizeof(*cmd);
> +	cmd = calloc(1, size);
> +	if (!cmd)
> +		return NULL;
> +
> +	cxl_cmd_ref(cmd);
> +	cmd->memdev = memdev;
> +
> +	return cmd;
> +}
> +
> +static int __do_cmd(struct cxl_cmd *cmd, int ioctl_cmd, int fd)
> +{
> +	void *cmd_buf;
> +	int rc;
> +
> +	switch (ioctl_cmd) {
> +	case CXL_MEM_QUERY_COMMANDS:
> +		cmd_buf = cmd->query_cmd;
> +		break;
> +	case CXL_MEM_SEND_COMMAND:
> +		cmd_buf = cmd->send_cmd;
> +		break;
> +	default:
> +		return -EINVAL;

__builtin_unreachable? I don't see other uses in this project, so perhaps no..

> +	}
> +
> +	rc = ioctl(fd, ioctl_cmd, cmd_buf);
> +	if (rc < 0)
> +		rc = -errno;
> +
> +	return rc;
> +}
> +
> +static int do_cmd(struct cxl_cmd *cmd, int ioctl_cmd)
> +{
> +	char *path;
> +	struct stat st;
> +	unsigned int major, minor;
> +	int rc = 0, fd;
> +	struct cxl_memdev *memdev = cmd->memdev;
> +	struct cxl_ctx *ctx = cxl_memdev_get_ctx(memdev);
> +	const char *devname = cxl_memdev_get_devname(memdev);
> +
> +	major = cxl_memdev_get_major(memdev);
> +	minor = cxl_memdev_get_minor(memdev);
> +
> +	if (asprintf(&path, "/dev/cxl/%s", devname) < 0)
> +		return -ENOMEM;
> +
> +	fd = open(path, O_RDWR);
> +	if (fd < 0) {
> +		err(ctx, "failed to open %s: %s\n", path, strerror(errno));
> +		rc = -errno;
> +		goto out;
> +	}
> +
> +	if (fstat(fd, &st) >= 0 && S_ISCHR(st.st_mode)
> +			&& major(st.st_rdev) == major
> +			&& minor(st.st_rdev) == minor) {
> +		rc = __do_cmd(cmd, ioctl_cmd, fd);

I don't see anything wrong with this, but could you explain a case where
major/minor would not be the right thing? Is this just like super paranoid
guarding?

I mean, if you ended up with the wrong device types, your IOCTLs will fail
anyway.

> +	} else {
> +		err(ctx, "failed to validate %s as a CXL memdev node\n", path);
> +		rc = -ENXIO;
> +	}
> +	close(fd);
> +out:
> +	free(path);
> +	return rc;
> +}
> +
> +static int alloc_do_query(struct cxl_cmd *cmd, int num_cmds)
> +{
> +	struct cxl_ctx *ctx = cxl_memdev_get_ctx(cmd->memdev);
> +	int rc;
> +
> +	rc = cxl_cmd_alloc_query(cmd, num_cmds);
> +	if (rc)
> +		return rc;
> +
> +	rc = do_cmd(cmd, CXL_MEM_QUERY_COMMANDS);
> +	if (rc < 0)
> +		err(ctx, "%s: query commands failed: %s\n",
> +			cxl_memdev_get_devname(cmd->memdev),
> +			strerror(-rc));
> +	return rc;
> +}
> +
> +static int cxl_cmd_do_query(struct cxl_cmd *cmd)
> +{
> +	struct cxl_memdev *memdev = cmd->memdev;
> +	struct cxl_ctx *ctx = cxl_memdev_get_ctx(memdev);
> +	const char *devname = cxl_memdev_get_devname(memdev);
> +	int rc, n_commands;
> +
> +	switch (cmd->query_status) {
> +	case CXL_CMD_QUERY_OK:
> +		return 0;
> +	case CXL_CMD_QUERY_UNSUPPORTED:
> +		return -EOPNOTSUPP;

This makes me a little bit nervous. I've been leaving open the possibility that
command support might be transient. For instance, perhaps at boot a command is
not supported, but you change some device state and then it later becomes
supported. Or perhaps you have a command available and you go into some sort of
degraded performance mode which removes a set of commands.

Thoughts?

> +	case CXL_CMD_QUERY_NOT_RUN:
> +		break;
> +	default:
> +		err(ctx, "%s: Unknown query_status %d\n",
> +			devname, cmd->query_status);
> +		return -EINVAL;

Same about unreachable...

> +	}
> +
> +	rc = alloc_do_query(cmd, 0);
> +	if (rc)
> +		return rc;
> +
> +	n_commands = cmd->query_cmd->n_commands;
> +	dbg(ctx, "%s: supports %d commands\n", devname, n_commands);
> +
> +	return alloc_do_query(cmd, n_commands);
> +}
> +
> +static int cxl_cmd_validate(struct cxl_cmd *cmd, u32 cmd_id)
> +{
> +	struct cxl_memdev *memdev = cmd->memdev;
> +	struct cxl_mem_query_commands *query = cmd->query_cmd;
> +	const char *devname = cxl_memdev_get_devname(memdev);
> +	struct cxl_ctx *ctx = cxl_memdev_get_ctx(memdev);
> +	u32 i;
> +
> +	for (i = 0; i < query->n_commands; i++) {
> +		struct cxl_command_info *cinfo = &query->commands[i];
> +		const char *cmd_name = cxl_command_names[cinfo->id].name;
> +
> +		if (cinfo->id != cmd_id)
> +			continue;
> +
> +		dbg(ctx, "%s: %s: in: %d, out %d, flags: %#08x\n",
> +			devname, cmd_name, cinfo->size_in,
> +			cinfo->size_out, cinfo->flags);
> +
> +		if (cinfo->flags & CXL_MEM_COMMAND_FLAG_KERNEL) {
> +			err(ctx, "%s: %s: reserved for kernel use\n",
> +				devname, cmd_name);
> +			return -EPERM;
> +		}
> +		cmd->query_idx = i;
> +		cmd->query_status = CXL_CMD_QUERY_OK;
> +		return 0;
> +	}
> +	cmd->query_status = CXL_CMD_QUERY_UNSUPPORTED;
> +	return -EOPNOTSUPP;
> +}
> +
> +CXL_EXPORT int cxl_cmd_set_input_payload(struct cxl_cmd *cmd, void *buf,
> +		int size)
> +{
> +	struct cxl_memdev *memdev = cmd->memdev;
> +
> +	if (size > memdev->payload_max || size < 0)
> +		return -EINVAL;
> +
> +	if (!buf) {
> +
> +		/* If the user didn't supply a buffer, allocate it */
> +		cmd->input_payload = calloc(1, size);
> +		if (!cmd->input_payload)
> +			return -ENOMEM;
> +		cmd->send_cmd->in.payload = (u64)cmd->input_payload;
> +	} else {
> +		/*
> +		 * Use user-buffer as is. If an automatic allocation was
> +		 * previously made (based on a fixed size from query),
> +		 * it will get freed during unref.
> +		 */
> +		cmd->send_cmd->in.payload = (u64)buf;
> +	}
> +	cmd->send_cmd->in.size = size;
> +
> +	return 0;
> +}
> +
> +CXL_EXPORT int cxl_cmd_set_output_payload(struct cxl_cmd *cmd, void *buf,
> +		int size)
> +{
> +	struct cxl_memdev *memdev = cmd->memdev;
> +
> +	if (size > memdev->payload_max || size < 0)
> +		return -EINVAL;
> +
> +	if (!buf) {
> +
> +		/* If the user didn't supply a buffer, allocate it */
> +		cmd->output_payload = calloc(1, size);
> +		if (!cmd->output_payload)
> +			return -ENOMEM;
> +		cmd->send_cmd->out.payload = (u64)cmd->output_payload;
> +	} else {
> +		/*
> +		 * Use user-buffer as is. If an automatic allocation was
> +		 * previously made (based on a fixed size from query),
> +		 * it will get freed during unref.
> +		 */
> +		cmd->send_cmd->out.payload = (u64)buf;
> +	}
> +	cmd->send_cmd->out.size = size;
> +
> +	return 0;
> +}
> +
> +static int cxl_cmd_alloc_send(struct cxl_cmd *cmd, u32 cmd_id)
> +{
> +	struct cxl_mem_query_commands *query = cmd->query_cmd;
> +	struct cxl_command_info *cinfo = &query->commands[cmd->query_idx];
> +	size_t size;
> +
> +	if (!query)
> +		return -EINVAL;
> +
> +	size = sizeof(struct cxl_send_command);
> +	cmd->send_cmd = calloc(1, size);
> +	if (!cmd->send_cmd)
> +		return -ENOMEM;
> +
> +	if (cinfo->id != cmd_id)
> +		return -EINVAL;
> +
> +	cmd->send_cmd->id = cmd_id;
> +
> +	if (cinfo->size_in > 0) {
> +		cmd->input_payload = calloc(1, cinfo->size_in);
> +		if (!cmd->input_payload)
> +			return -ENOMEM;
> +		cmd->send_cmd->in.payload = (u64)cmd->input_payload;
> +		cmd->send_cmd->in.size = cinfo->size_in;
> +	}
> +	if (cinfo->size_out > 0) {
> +		cmd->output_payload = calloc(1, cinfo->size_out);
> +		if (!cmd->output_payload)
> +			return -ENOMEM;
> +		cmd->send_cmd->out.payload = (u64)cmd->output_payload;
> +		cmd->send_cmd->out.size = cinfo->size_out;
> +	}
> +
> +	return 0;
> +}
> +
> +static struct cxl_cmd *cxl_cmd_new_generic(struct cxl_memdev *memdev,
> +		u32 cmd_id)
> +{
> +	const char *devname = cxl_memdev_get_devname(memdev);
> +	struct cxl_ctx *ctx = cxl_memdev_get_ctx(memdev);
> +	struct cxl_cmd *cmd;
> +	int rc;
> +
> +	cmd = cxl_cmd_new(memdev);
> +	if (!cmd)
> +		return NULL;
> +
> +	rc = cxl_cmd_do_query(cmd);
> +	if (rc) {
> +		err(ctx, "%s: query returned: %s\n", devname, strerror(-rc));
> +		goto fail;
> +	}
> +
> +	rc = cxl_cmd_validate(cmd, cmd_id);
> +	if (rc) {
> +		errno = -rc;
> +		goto fail;
> +	}
> +
> +	rc = cxl_cmd_alloc_send(cmd, cmd_id);
> +	if (rc) {
> +		errno = -rc;
> +		goto fail;
> +	}
> +
> +	return cmd;
> +
> +fail:
> +	cxl_cmd_unref(cmd);
> +	return NULL;
> +}
> +
> +CXL_EXPORT const char *cxl_cmd_get_devname(struct cxl_cmd *cmd)
> +{
> +	return cxl_memdev_get_devname(cmd->memdev);
> +}
> +
> +CXL_EXPORT struct cxl_cmd *cxl_cmd_new_raw(struct cxl_memdev *memdev,
> +		int opcode)
> +{
> +	struct cxl_cmd *cmd;
> +
> +	/* opcode '0' is reserved */
> +	if (opcode <= 0) {
> +		errno = EINVAL;
> +		return NULL;
> +	}
> +
> +	cmd = cxl_cmd_new_generic(memdev, CXL_MEM_COMMAND_ID_RAW);
> +	if (!cmd)
> +		return NULL;
> +
> +	cmd->send_cmd->raw.opcode = opcode;
> +	return cmd;
> +}
> +
> +CXL_EXPORT int cxl_cmd_submit(struct cxl_cmd *cmd)
> +{
> +	struct cxl_memdev *memdev = cmd->memdev;
> +	const char *devname = cxl_memdev_get_devname(memdev);
> +	struct cxl_ctx *ctx = cxl_memdev_get_ctx(memdev);
> +	int rc;
> +
> +	switch (cmd->query_status) {
> +	case CXL_CMD_QUERY_OK:
> +		break;
> +	case CXL_CMD_QUERY_UNSUPPORTED:
> +		return -EOPNOTSUPP;
> +	case CXL_CMD_QUERY_NOT_RUN:
> +		return -EINVAL;
> +	default:
> +		err(ctx, "%s: Unknown query_status %d\n",
> +			devname, cmd->query_status);
> +		return -EINVAL;
> +	}
> +
> +	dbg(ctx, "%s: submitting SEND cmd: in: %d, out: %d\n", devname,
> +		cmd->send_cmd->in.size, cmd->send_cmd->out.size);
> +	rc = do_cmd(cmd, CXL_MEM_SEND_COMMAND);
> +	if (rc < 0)
> +		err(ctx, "%s: send command failed: %s\n",
> +			devname, strerror(-rc));
> +	cmd->status = cmd->send_cmd->retval;
> +	dbg(ctx, "%s: got SEND cmd: in: %d, out: %d, retval: %d\n", devname,
> +		cmd->send_cmd->in.size, cmd->send_cmd->out.size, cmd->status);
> +
> +	return rc;
> +}
> +
> +CXL_EXPORT int cxl_cmd_get_mbox_status(struct cxl_cmd *cmd)
> +{
> +	return cmd->status;
> +}
> +
> +CXL_EXPORT int cxl_cmd_get_out_size(struct cxl_cmd *cmd)
> +{
> +	return cmd->send_cmd->out.size;
> +}
> diff --git a/cxl/libcxl.h b/cxl/libcxl.h
> index fd06790..6e87b80 100644
> --- a/cxl/libcxl.h
> +++ b/cxl/libcxl.h
> @@ -48,6 +48,17 @@ const char *cxl_memdev_get_firmware_verison(struct cxl_memdev *memdev);
>               memdev != NULL; \
>               memdev = cxl_memdev_get_next(memdev))
>  
> +struct cxl_cmd;
> +const char *cxl_cmd_get_devname(struct cxl_cmd *cmd);
> +struct cxl_cmd *cxl_cmd_new_raw(struct cxl_memdev *memdev, int opcode);
> +int cxl_cmd_set_input_payload(struct cxl_cmd *cmd, void *in, int size);
> +int cxl_cmd_set_output_payload(struct cxl_cmd *cmd, void *out, int size);
> +void cxl_cmd_ref(struct cxl_cmd *cmd);
> +void cxl_cmd_unref(struct cxl_cmd *cmd);
> +int cxl_cmd_submit(struct cxl_cmd *cmd);
> +int cxl_cmd_get_mbox_status(struct cxl_cmd *cmd);
> +int cxl_cmd_get_out_size(struct cxl_cmd *cmd);
> +
>  #ifdef __cplusplus
>  } /* extern "C" */
>  #endif
> diff --git a/cxl/lib/libcxl.sym b/cxl/lib/libcxl.sym
> index 0f6ecad..493429c 100644
> --- a/cxl/lib/libcxl.sym
> +++ b/cxl/lib/libcxl.sym
> @@ -27,3 +27,16 @@ global:
>  	cxl_memdev_get_ram_size;
>  	cxl_memdev_get_firmware_verison;
>  } LIBCXL_1;
> +
> +LIBCXL_3 {
> +global:
> +	cxl_cmd_get_devname;
> +	cxl_cmd_new_raw;
> +	cxl_cmd_set_input_payload;
> +	cxl_cmd_set_output_payload;
> +	cxl_cmd_ref;
> +	cxl_cmd_unref;
> +	cxl_cmd_submit;
> +	cxl_cmd_get_mbox_status;
> +	cxl_cmd_get_out_size;
> +} LIBCXL_2;
> -- 
> 2.29.2
> 
_______________________________________________
Linux-nvdimm mailing list -- linux-nvdimm@lists.01.org
To unsubscribe send an email to linux-nvdimm-leave@lists.01.org

  reply	other threads:[~2021-02-22 21:55 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-02-19  2:03 [ndctl PATCH v2 00/13] Initial CXL support Vishal Verma
2021-02-19  2:03 ` [ndctl PATCH v2 01/13] cxl: add a cxl utility and libcxl library Vishal Verma
2021-02-22 21:36   ` Ben Widawsky
2021-02-23 19:23     ` Verma, Vishal L
2021-02-19  2:03 ` [ndctl PATCH v2 02/13] cxl: add a local copy of the cxl_mem UAPI header Vishal Verma
2021-02-22 21:41   ` Ben Widawsky
2021-02-19  2:03 ` [ndctl PATCH v2 03/13] libcxl: add support for command query and submission Vishal Verma
2021-02-22 21:55   ` Ben Widawsky [this message]
2021-02-19  2:03 ` [ndctl PATCH v2 04/13] libcxl: add support for the 'Identify Device' command Vishal Verma
2021-02-22 22:02   ` Ben Widawsky
2021-02-19  2:03 ` [ndctl PATCH v2 05/13] test: rename 'ndctl_test' to 'test_ctx' Vishal Verma
2021-02-19  2:03 ` [ndctl PATCH v2 06/13] test: rename 'ndctl_test_*' helpers to 'test_*' Vishal Verma
2021-02-19  2:03 ` [ndctl PATCH v2 07/13] test: introduce a libcxl unit test Vishal Verma
2021-02-22 22:15   ` Ben Widawsky
2021-02-19  2:03 ` [ndctl PATCH v2 08/13] libcxl: add GET_HEALTH_INFO mailbox command and accessors Vishal Verma
2021-02-19  2:03 ` [ndctl PATCH v2 09/13] libcxl: add support for the 'GET_LSA' command Vishal Verma
2021-02-19  2:03 ` [ndctl PATCH v2 10/13] util/hexdump: Add a util helper to print a buffer in hex Vishal Verma
2021-02-19  2:03 ` [ndctl PATCH v2 11/13] test/libcxl: add a test for {set, get}_lsa commands Vishal Verma
2021-02-19  2:03 ` [ndctl PATCH v2 12/13] Documentation/cxl: add library API documentation Vishal Verma
2021-02-19  2:03 ` [ndctl PATCH v2 13/13] test/libcxl: introduce a command size fuzzing test Vishal Verma

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20210222215539.s3n6yj5lgmdohypt@intel.com \
    --to=ben.widawsky@intel.com \
    --cc=linux-cxl@vger.kernel.org \
    --cc=linux-nvdimm@lists.01.org \
    --cc=vishal.l.verma@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).