public inbox for passt-dev@passt.top
 help / color / mirror / code / Atom feed
From: David Gibson <david@gibson.dropbear.id.au>
To: Anshu Kumari <anskuma@redhat.com>
Cc: passt-dev@passt.top, sbrivio@redhat.com, jmaloy@redhat.com,
	lvivier@redhat.com
Subject: Re: [PATCH v4 2/4] dhcp: Add option overload
Date: Fri, 19 Jun 2026 13:39:25 +1000	[thread overview]
Message-ID: <ajS57XHwXBqsV2mS@zatzit> (raw)
In-Reply-To: <20260617132243.1499556-3-anskuma@redhat.com>

[-- Attachment #1: Type: text/plain, Size: 8126 bytes --]

On Wed, Jun 17, 2026 at 06:52:36PM +0530, Anshu Kumari wrote:
> When the options field is full, overflow remaining DHCP options into
> the sname and file fields per RFC 2132 option 52.
> 
> Per RFC 2132, Section 9.5, the boot file name is always placed in the
> 'file' header field.  When a boot file is set, the file field is
> reserved from overload and overflow uses only the sname field.
> 
> Link: https://bugs.passt.top/show_bug.cgi?id=192
> Signed-off-by: Anshu Kumari <anskuma@redhat.com>
> ---
> v4:
>   - Converted overload #defines to enum dhcp_overload.
>   - Fixed missing whitespace in comment before */.
>   - Boot file name always placed in 'file' header field per RFC 2132,
>     Section 9.5; file field reserved from overload when bootfile is
>     set; option 67 suppressed from options area.
> 
> v3:
>   - Added RFC 2132 Section 9.3 reference comment on overload
>     constants.
>   - Use ARRAY_SIZE(opts) instead of raw 255 in fill_overflow().
>   - Swapped overflow order: try sname (64 bytes) first, then file
>     (128 bytes) — better packing and keeps file field available for
>     boot file name.
>   - Removed '&' from &reply.file.
>   - Removed '+1' from memcpy — reply.file already zeroed.
>   - opt_set_dns_search() max_len: OPT_MAX - 3 instead of
>     sizeof(m->o).
> 
> v2:
>   - Added #define DHCP_OVERLOAD_FILE and #define DHCP_OVERLOAD_SNAME constants
>   - Added comment documenting space reservation: /* Reserve 3 bytes for option 52 */
>   - Fixed DNS search length: sizeof(m->o) only, not combined with file+sname
>   - Removed dhcp_boot references — reply.file copy now reads from opts[67]
>   - Used DHCP_OVERLOAD_FILE constant in reply.file guard
> ---
>  dhcp.c | 108 ++++++++++++++++++++++++++++++++++++++++++++++++++++-----
>  1 file changed, 100 insertions(+), 8 deletions(-)
> 
> diff --git a/dhcp.c b/dhcp.c
> index 8e3ffd6..78790d8 100644
> --- a/dhcp.c
> +++ b/dhcp.c
> @@ -160,14 +160,79 @@ static bool fill_one(uint8_t *buf, size_t size, int o, int *offset)
>  	return false;
>  }
>  
> +/* RFC 2132, Section 9.3 - Option Overload */
> +enum dhcp_overload {
> +	DHCP_OVERLOAD_NONE  = 0,
> +	DHCP_OVERLOAD_FILE  = 1,
> +	DHCP_OVERLOAD_SNAME = 2,
> +};
> +
>  /**
> - * fill() - Fill options in message
> - * @m:		Message to fill
> + * fill_overflow() - Fill remaining options into sname and file fields
> + * @m:			Msg whose sname/file fields may be used for overflow
> + * @has_bootfile:	If true, reserve the file field for the boot file name
> + *
> + * Try the smaller sname field first: small options go there, leaving
> + * the larger file field available for big options. When @has_bootfile
> + * is set, the file field is reserved for the boot file name per
> + * RFC 2132, Section 9.5 and is not used for option overflow.
> + *
> + * Return: option 52 overload value
> + */
> +static enum dhcp_overload fill_overflow(struct msg *m, bool has_bootfile)
> +{
> +	enum dhcp_overload overload = DHCP_OVERLOAD_NONE;
> +	int sname_off = 0, file_off = 0;
> +	int o;
> +
> +	for (o = 0; (size_t)o < ARRAY_SIZE(opts); o++) {
> +		if (opts[o].slen == -1 || opts[o].sent)
> +			continue;
> +		fill_one(m->sname, sizeof(m->sname) - 1, o, &sname_off);

Ah, I think I see now.  The earlier debug() messages were removed in
favour of the ones later, because we don't want spurious messages if
the option doesn't fit in the main area but does fit in the overload.

That's a reasonable change, but needs to be better explained.  It also
all belongs in this patch, so the bits that leaked into the previous
patch should be here instead.

> +	}
> +
> +	if (!has_bootfile) {
> +		for (o = 0; (size_t)o < ARRAY_SIZE(opts); o++) {
> +			if (opts[o].slen == -1 || opts[o].sent)
> +				continue;
> +			if (fill_one(m->file, sizeof(m->file) - 1, o,
> +				     &file_off))
> +				debug("DHCP: skipping option %i"
> +				      " (overload full)", o);
> +		}
> +	} else {

Rather that checking for fill_one() failures only on the last pass,
then having this fallback case, you can have a single unconditional
pass at the end looking for skipped options.  Then fill_one()s
signature can be changed to void to match.

> +		for (o = 0; (size_t)o < ARRAY_SIZE(opts); o++) {
> +			if (opts[o].slen == -1 || opts[o].sent)
> +				continue;
> +			debug("DHCP: skipping option %i (overload full)", o);
> +		}
> +	}
> +
> +	if (sname_off) {
> +		m->sname[sname_off] = 255;
> +		overload |= DHCP_OVERLOAD_SNAME;
> +	}
> +
> +	if (file_off) {
> +		m->file[file_off] = 255;
> +		overload |= DHCP_OVERLOAD_FILE;
> +	}
> +
> +	return overload;
> +}
> +
> +/**
> + * fill() - Fill options in message, with overload into file/sname if needed
> + * @m:			Message to fill
> + * @overload:		Set to option 52 value (0 if none, 1/2/3 per RFC 2132)
> + * @has_bootfile:	Reserve file field for boot file name
>   *
>   * Return: current size of options field
>   */
> -static int fill(struct msg *m)
> +static int fill(struct msg *m, enum dhcp_overload *overload, bool has_bootfile)
>  {
> +	/* Reserve 3 bytes for option 52 (overload) if needed */
> +	size_t size = OPT_MAX - 3;
>  	int i, o, offset = 0;
>  
>  	for (o = 0; o < 255; o++)
> @@ -178,17 +243,25 @@ static int fill(struct msg *m)
>  	 * Put it there explicitly, unless requested via option 55.
>  	 */
>  	if (opts[55].clen > 0 && !memchr(opts[55].c, 53, opts[55].clen))
> -		fill_one(m->o, OPT_MAX, 53, &offset);
> +		fill_one(m->o, size, 53, &offset);
>  
>  	for (i = 0; i < opts[55].clen; i++) {
>  		o = opts[55].c[i];
>  		if (opts[o].slen != -1)
> -			fill_one(m->o, OPT_MAX, o, &offset);
> +			fill_one(m->o, size, o, &offset);
>  	}
>  
>  	for (o = 0; o < 255; o++) {
>  		if (opts[o].slen != -1 && !opts[o].sent)
> -			fill_one(m->o, OPT_MAX, o, &offset);
> +			fill_one(m->o, size, o, &offset);
> +	}
> +
> +	*overload = fill_overflow(m, has_bootfile);

Is there a particular reason to put fill_overflow() in its own
function, rather than just inline here?

> +
> +	if (*overload) {
> +		m->o[offset++] = 52;
> +		m->o[offset++] = 1;
> +		m->o[offset++] = *overload;
>  	}
>  
>  	m->o[offset++] = 255;
> @@ -301,6 +374,7 @@ static void opt_set_dns_search(const struct ctx *c, size_t max_len)
>  int dhcp(const struct ctx *c, struct iov_tail *data)
>  {
>  	char macstr[ETH_ADDRSTRLEN];
> +	enum dhcp_overload overload;
>  	size_t mlen, dlen, opt_len;
>  	struct in_addr mask, dst;
>  	struct ethhdr eh_storage;
> @@ -309,9 +383,12 @@ int dhcp(const struct ctx *c, struct iov_tail *data)
>  	const struct ethhdr *eh;
>  	const struct iphdr *iph;
>  	const struct udphdr *uh;
> +	uint8_t bootfile[128];
>  	struct msg m_storage;
>  	struct msg const *m;
> +	bool has_bootfile;
>  	struct msg reply;
> +	int bootfile_len;
>  	unsigned int i;
>  
>  	eh = IOV_REMOVE_HEADER(data, eh_storage);
> @@ -462,9 +539,24 @@ int dhcp(const struct ctx *c, struct iov_tail *data)
>  	}
>  
>  	if (!c->no_dhcp_dns_search)
> -		opt_set_dns_search(c, sizeof(m->o));
> +		opt_set_dns_search(c, OPT_MAX - 3);
> +
> +	/* RFC 2132, Section 9.5: put boot file name in the 'file' header
> +	 * field.  Suppress option 67 from the options area and reserve
> +	 * the file field from overload.
> +	 */
> +	has_bootfile = opts[67].slen > 0 &&
> +		       (size_t)opts[67].slen < sizeof(reply.file);
> +	if (has_bootfile) {
> +		memcpy(bootfile, opts[67].s, opts[67].slen);
> +		bootfile_len = opts[67].slen;
> +		opts[67].slen = -1;
> +	}
> +
> +	dlen = offsetof(struct msg, o) + fill(&reply, &overload, has_bootfile);
>  
> -	dlen = offsetof(struct msg, o) + fill(&reply);
> +	if (has_bootfile)
> +		memcpy(reply.file, bootfile, bootfile_len);
>  
>  	if (m->flags & FLAG_BROADCAST)
>  		dst = in4addr_broadcast;
> -- 
> 2.54.0
> 

-- 
David Gibson (he or they)	| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you, not the other way
				| around.
http://www.ozlabs.org/~dgibson

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

  reply	other threads:[~2026-06-19  3:54 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-06-17 13:22 [PATCH v4 0/4] Add --dhcp-boot and --dhcp-opt options Anshu Kumari
2026-06-17 13:22 ` [PATCH v4 1/4] dhcp: Refactor fill_one() to operate on a generic buffer Anshu Kumari
2026-06-19  3:26   ` David Gibson
2026-06-17 13:22 ` [PATCH v4 2/4] dhcp: Add option overload Anshu Kumari
2026-06-19  3:39   ` David Gibson [this message]
2026-06-19  7:55     ` Anshu Kumari
2026-06-17 13:22 ` [PATCH v4 3/4] dhcp: Add --dhcp-opt with option table and value parser Anshu Kumari
2026-06-19  3:52   ` David Gibson
2026-06-19  7:29     ` Anshu Kumari
2026-06-17 13:22 ` [PATCH v4 4/4] conf: Add --dhcp-boot command-line option Anshu Kumari
2026-06-19  3:53   ` David Gibson

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=ajS57XHwXBqsV2mS@zatzit \
    --to=david@gibson.dropbear.id.au \
    --cc=anskuma@redhat.com \
    --cc=jmaloy@redhat.com \
    --cc=lvivier@redhat.com \
    --cc=passt-dev@passt.top \
    --cc=sbrivio@redhat.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.
Code repositories for project(s) associated with this public inbox

	https://passt.top/passt

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 IMAP folder(s).