netdev - Re: [PATCH ethtool-next 2/5] pause: add --json support

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 23 Sep 2020 17:10:30 -0700
From:   Jacob Keller <jacob.e.keller@...el.com>
To:     Jakub Kicinski <kuba@...nel.org>, mkubecek@...e.cz
Cc:     netdev@...r.kernel.org
Subject: Re: [PATCH ethtool-next 2/5] pause: add --json support



On 9/15/2020 4:52 PM, Jakub Kicinski wrote:
> No change in normal text output:
> 
>  # ./ethtool  -a eth0
> Pause parameters for eth0:
> Autonegotiate:	on
> RX:		on
> TX:		on
> RX negotiated: on
> TX negotiated: on
> 
> JSON:
> 
>  # ./ethtool --json -a eth0
> [ {
>         "ifname": "eth0",
>         "autonegotiate": true,
>         "rx": true,
>         "tx": true,
>         "negotiated": {
>             "rx": true,
>             "tx": true
>         }
>     } ]


Makes sense, we report all of these as a single dictionary with
key-values. Putting negotiated as a subsection seems reasonable as well.
I guess we could split this so that tx and rx have their own section.
but I think this is good.

This looks good to me.

Reviewed-by: Jacob Keller <jacob.e.keller@...il.com>

> 
> Signed-off-by: Jakub Kicinski <kuba@...nel.org>
> ---
>  netlink/netlink.h | 12 ++++++++++--
>  netlink/pause.c   | 45 +++++++++++++++++++++++++++++++++++----------
>  2 files changed, 45 insertions(+), 12 deletions(-)
> 
> diff --git a/netlink/netlink.h b/netlink/netlink.h
> index dd4a02bcc916..4916d25ed5c0 100644
> --- a/netlink/netlink.h
> +++ b/netlink/netlink.h
> @@ -106,9 +106,17 @@ static inline const char *u8_to_bool(const struct nlattr *attr)
>  		return "n/a";
>  }
>  
> -static inline void show_bool(const struct nlattr *attr, const char *label)
> +static inline void show_bool(const char *key, const char *fmt,
> +			     const struct nlattr *attr)
>  {
> -	printf("%s%s\n", label, u8_to_bool(attr));
> +	if (is_json_context()) {
> +		if (attr) {
> +			print_bool(PRINT_JSON, key, NULL,
> +				   mnl_attr_get_u8(attr));
> +		}
> +	} else {
> +		print_string(PRINT_FP, NULL, fmt, u8_to_bool(attr));

Ok. The change from label to fmt makes sense because we now pass a
format string instead of a label, and we use print_string instead of printf.

> +	}
>  }
>  
>  /* misc */
> diff --git a/netlink/pause.c b/netlink/pause.c
> index 7b6b3a1d2c10..30ecdccb15eb 100644
> --- a/netlink/pause.c
> +++ b/netlink/pause.c
> @@ -72,8 +72,16 @@ static int pause_autoneg_cb(const struct nlmsghdr *nlhdr, void *data)
>  		else if (peer.pause)
>  			tx_status = true;
>  	}
> -	printf("RX negotiated: %s\nTX negotiated: %s\n",
> -	       rx_status ? "on" : "off", tx_status ? "on" : "off");
> +
> +	if (is_json_context()) {
> +		open_json_object("negotiated");
> +		print_bool(PRINT_JSON, "rx", NULL, rx_status);
> +		print_bool(PRINT_JSON, "tx", NULL, tx_status);
> +		close_json_object();
> +	} else {
> +		printf("RX negotiated: %s\nTX negotiated: %s\n",
> +		       rx_status ? "on" : "off", tx_status ? "on" : "off");
> +	}
>  

Why not use print_string here like show_bool did?

>  	return MNL_CB_OK;
>  }
> @@ -121,21 +129,34 @@ int pause_reply_cb(const struct nlmsghdr *nlhdr, void *data)
>  		return err_ret;
>  
>  	if (silent)
> -		putchar('\n');
> -	printf("Pause parameters for %s:\n", nlctx->devname);
> -	show_bool(tb[ETHTOOL_A_PAUSE_AUTONEG], "Autonegotiate:\t");
> -	show_bool(tb[ETHTOOL_A_PAUSE_RX], "RX:\t\t");
> -	show_bool(tb[ETHTOOL_A_PAUSE_TX], "TX:\t\t");
> +		print_nl();
> +
> +	open_json_object(NULL);
> +
> +	print_string(PRINT_ANY, "ifname", "Pause parameters for %s:\n",
> +		     nlctx->devname);
> +

Ah, nice, so this prints based on the current format and will just print
the ifname instead of showing the format string. Neat.

> +	show_bool("autonegotiate", "Autonegotiate:\t%s\n",
> +		  tb[ETHTOOL_A_PAUSE_AUTONEG]);
> +	show_bool("rx", "RX:\t\t%s\n", tb[ETHTOOL_A_PAUSE_RX]);
> +	show_bool("tx", "TX:\t\t%s\n", tb[ETHTOOL_A_PAUSE_TX]);
> +
>  	if (!nlctx->is_monitor && tb[ETHTOOL_A_PAUSE_AUTONEG] &&
>  	    mnl_attr_get_u8(tb[ETHTOOL_A_PAUSE_AUTONEG])) {
>  		ret = show_pause_autoneg_status(nlctx);
>  		if (ret < 0)
> -			return err_ret;
> +			goto err_close_dev;
>  	}
>  	if (!silent)
> -		putchar('\n');
> +		print_nl();
> +
> +	close_json_object();
>  
>  	return MNL_CB_OK;
> +
> +err_close_dev:
> +	close_json_object();
> +	return err_ret;
>  }
>  
>  int nl_gpause(struct cmd_context *ctx)
> @@ -156,7 +177,11 @@ int nl_gpause(struct cmd_context *ctx)
>  				      ETHTOOL_A_PAUSE_HEADER, 0);
>  	if (ret < 0)
>  		return ret;
> -	return nlsock_send_get_request(nlsk, pause_reply_cb);
> +
> +	new_json_obj(ctx->json);
> +	ret = nlsock_send_get_request(nlsk, pause_reply_cb);
> +	delete_json_obj();
> +	return ret;
>  }
>  
>  /* PAUSE_SET */
>