From mboxrd@z Thu Jan 1 00:00:00 1970 Authentication-Results: passt.top; dmarc=none (p=none dis=none) header.from=gibson.dropbear.id.au Authentication-Results: passt.top; dkim=pass (2048-bit key; secure) header.d=gibson.dropbear.id.au header.i=@gibson.dropbear.id.au header.a=rsa-sha256 header.s=202502 header.b=a/8BLEFt; dkim-atps=neutral Received: from mail.ozlabs.org (mail.ozlabs.org [IPv6:2404:9400:2221:ea00::3]) by passt.top (Postfix) with ESMTPS id 74BD45A061F for ; Thu, 03 Apr 2025 07:38:18 +0200 (CEST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gibson.dropbear.id.au; s=202502; t=1743658684; bh=LLs/Tr+Cd231p1SC04CJ9D+Km+uNQ2x9wVL3jcoseG0=; h=Date:From:To:Cc:Subject:References:In-Reply-To:From; b=a/8BLEFt8jwqSPWPUz197ed8YBHY17tXwgwVb2fDu0705+WBbRUsXtGJszfIuEHCM hcaPvuy+uyWg6J2iXXFHy3xbF79FhR1XCCREKF9Hr55EIG6x0wxIweHcKXOuj3OLI+ AlAWgT2qxLs0pablGXq3PijWtaHqV0DBrk/eGej12ZlLApK2v0pRguGsSVmy7X2ngz XIwY0xJYB8lWkuJMxwXoIc+eAa46OIufGt35k08bEVMAIeYTXAABUqBXgGQ4SwaUHS J5xmfejLqAx+/XhF7Mr0WfXpha3SwmzzYTtmzQTNcu25bn6jKx2DUnQS2XteDOhosu UbUMjJDGZH49A== Received: by gandalf.ozlabs.org (Postfix, from userid 1007) id 4ZSr9m5f2Kz4xN4; Thu, 3 Apr 2025 16:38:04 +1100 (AEDT) Date: Thu, 3 Apr 2025 16:20:22 +1100 From: David Gibson To: Laurent Vivier Subject: Re: [PATCH 11/18] tcp: Convert tcp_data_from_tap() to use iov_tail Message-ID: References: <20250402172343.858187-1-lvivier@redhat.com> <20250402172343.858187-12-lvivier@redhat.com> MIME-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha256; protocol="application/pgp-signature"; boundary="lF+VV0GX4En1vvkj" Content-Disposition: inline In-Reply-To: <20250402172343.858187-12-lvivier@redhat.com> Message-ID-Hash: BODPSINW63ENPRJNYDW65JHGMMC5F3Y4 X-Message-ID-Hash: BODPSINW63ENPRJNYDW65JHGMMC5F3Y4 X-MailFrom: dgibson@gandalf.ozlabs.org X-Mailman-Rule-Misses: dmarc-mitigation; no-senders; approved; emergency; loop; banned-address; member-moderation; nonmember-moderation; administrivia; implicit-dest; max-recipients; max-size; news-moderation; no-subject; digests; suspicious-header CC: passt-dev@passt.top X-Mailman-Version: 3.3.8 Precedence: list List-Id: Development discussion and patches for passt Archived-At: Archived-At: List-Archive: List-Archive: List-Help: List-Owner: List-Post: List-Subscribe: List-Unsubscribe: --lF+VV0GX4En1vvkj Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Wed, Apr 02, 2025 at 07:23:36PM +0200, Laurent Vivier wrote: > Use packet_base() and extract headers using IOV_PEEK_HEADER() > rather than packet_get(). >=20 > Signed-off-by: Laurent Vivier > --- > tcp.c | 24 ++++++++++++++---------- > 1 file changed, 14 insertions(+), 10 deletions(-) >=20 > diff --git a/tcp.c b/tcp.c > index 790714a08793..a9c04551d9d6 100644 > --- a/tcp.c > +++ b/tcp.c > @@ -1643,14 +1643,19 @@ static int tcp_data_from_tap(const struct ctx *c,= struct tcp_tap_conn *conn, > for (i =3D idx, iov_i =3D 0; i < (int)p->count; i++) { > uint32_t seq, seq_offset, ack_seq; > const struct tcphdr *th; > - char *data; > + struct iov_tail data; > + unsigned int count; > + struct tcphdr thc; > size_t off; > =20 > - th =3D packet_get(p, i, 0, sizeof(*th), &len); > + if (!packet_base(p, i, &data)) > + return -1; > + > + th =3D IOV_PEEK_HEADER(&data, thc); > if (!th) > return -1; > - len +=3D sizeof(*th); > =20 > + len =3D iov_tail_size(&data); > off =3D th->doff * 4UL; > if (off < sizeof(*th) || off > len) > return -1; > @@ -1661,9 +1666,7 @@ static int tcp_data_from_tap(const struct ctx *c, s= truct tcp_tap_conn *conn, > } > =20 > len -=3D off; > - data =3D packet_get(p, i, off, len, NULL); > - if (!data) > - continue; > + data.off =3D off; You can use iov_drop() rather than reaching into the internals of iov_tail here, no? > seq =3D ntohl(th->seq); > if (SEQ_LT(seq, conn->seq_from_tap) && len <=3D 1) { > @@ -1737,10 +1740,11 @@ static int tcp_data_from_tap(const struct ctx *c,= struct tcp_tap_conn *conn, > continue; > } > =20 > - tcp_iov[iov_i].iov_base =3D data + seq_offset; > - tcp_iov[iov_i].iov_len =3D len - seq_offset; > - seq_from_tap +=3D tcp_iov[iov_i].iov_len; > - iov_i++; > + count =3D iov_copy(&tcp_iov[iov_i], UIO_MAXIOV - iov_i, > + &data.iov[0], data.cnt, data.off + seq_offset, > + len - seq_offset); Here again it matters if you run out of space in the destination iov, and I don't think you have a check above which prevents it. > + seq_from_tap +=3D iov_size(&tcp_iov[iov_i], count); We already called iov_size() on &data above. We should be able to derive the total length here from that minus headers, without having to recount the IOV, no? > + iov_i +=3D count; > =20 > if (keep =3D=3D i) > keep =3D -1; --=20 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 --lF+VV0GX4En1vvkj Content-Type: application/pgp-signature; name=signature.asc -----BEGIN PGP SIGNATURE----- iQIzBAEBCAAdFiEEO+dNsU4E3yXUXRK2zQJF27ox2GcFAmfuGpUACgkQzQJF27ox 2GfG3A/+OMaObJZwX3/umZfSCvZohqVknjOxGuTsEQlAxbhMD0AxACKigo/9HQ42 3q/0Xr6+BjzbOIf09OZD6kMfQhDsM3rVGMzzDiNfLQN+RZ+nzqNA5L3YOirGyjUZ 6cwHj2eE4IjpjGDcYT7JyQ7N4iZ+5S36fZfMO5reZyw3NHQUJRiV2Bn9HIoqZ5aN zjCRxJX+M2S4VuKjXM3bq7OntXRHk3jEV3GrfiXcfVEOoKn9/JOVeRGFmr7W5Io3 TE08UXDdHsdXN1Xd8iUpPuK7AZh9v/MXErGQbqvIjYn41LISDwEd7SNtQ0H+TutE UysRmBdmUWo47Xp0pCb7yTqZbrhVBQ9S1HqDxe+AeKfbKOXJ/PIpCw8FkuzNbzjZ m7CHtSKLzRVDbg4n5QK2gxpLnKfhXwKf3K3CN7aB7+/Yl+bjFDuybSdb4iC+NIYz iu04R84ha8wUsIezN7ULDk2ZwyBlFHYGf68UCwRk8xJdQcJxP5NAM4EYsBXj50nh AaayrORWTphmaTlHYYMpJYABNJ/DBR3fmzzKCe4fOQrH/wnI02I4QF7aEY9Nm206 Q/TPD4iz7THcC0NKCEv1ZFp8t8YakXK4j8KfhGEY86ttah4dxL5+LHt/msZz5ZRP x5KsIqYmX4+fVwq2sateAjAbgrRYOxGuzsVfeaypwVA42UcyXxA= =NXkk -----END PGP SIGNATURE----- --lF+VV0GX4En1vvkj--