From mboxrd@z Thu Jan 1 00:00:00 1970 Authentication-Results: passt.top; dmarc=pass (p=quarantine dis=none) header.from=redhat.com Authentication-Results: passt.top; dkim=pass (1024-bit key; unprotected) header.d=redhat.com header.i=@redhat.com header.a=rsa-sha256 header.s=mimecast20190719 header.b=elt6AQdT; dkim-atps=neutral Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) by passt.top (Postfix) with ESMTPS id DDA6B5A0619 for ; Thu, 30 Oct 2025 09:26:10 +0100 (CET) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1761812769; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=udf2DD2sg3PCGkiNUJQekFcqGUTWmAeUikIDqTiNbQo=; b=elt6AQdTAkhM8eONhiihITHYsn0/HjQtTsHvSd7VtQ0+RpZg6Esy7qmSHzP98SEdEcs0wD /OL2f3ApkTLjpiX8c6SJuCUsLBTcxkPAb+Xm1U6jKYF01FUaXdE2UGxTVePRtBVg79DDcF VWWzPNaLy0gUElfnS8RslN4XTP8onc0= Received: from mail-ed1-f70.google.com (mail-ed1-f70.google.com [209.85.208.70]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-187-o02iXesfNHipoTrjrpIGOw-1; Thu, 30 Oct 2025 04:26:07 -0400 X-MC-Unique: o02iXesfNHipoTrjrpIGOw-1 X-Mimecast-MFC-AGG-ID: o02iXesfNHipoTrjrpIGOw_1761812766 Received: by mail-ed1-f70.google.com with SMTP id 4fb4d7f45d1cf-63e0fe97842so1229645a12.3 for ; Thu, 30 Oct 2025 01:26:07 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1761812766; x=1762417566; h=content-transfer-encoding:cc:to:subject:message-id:date:from :in-reply-to:references:mime-version:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=udf2DD2sg3PCGkiNUJQekFcqGUTWmAeUikIDqTiNbQo=; b=Eq19Eoe+kcjdWmO1NW7OxwdiXsfei0USwirP66l9ZXIpXzHef95jgTdhAUg3thfo9B qLuCe0HJa06Tpxh18c9Z6CpGcJpIZbRElQf2vPzKqUVzzju4ay5gOQ3iy2DqXDaeR4Du o/HKwsfJHN3TIu8F40XRGdo8wnuyfqos2yiJkhPojY0rRbzCc4DnAUtOBA2Bc4hauTyS LHqCJkSE8Kom3+8TXHQcU5T5Al113m82bUn3WAO8fIfVMsW2E5yKsmKTBmC3E7/4Fo0J T1IuHpwEtDEGx1XjGQro3596+5s/g/hv3OJXJZiQ3uaxqhBg5SdaY8q60NSteXP07dUN LMFQ== X-Gm-Message-State: AOJu0YyX86LLihR0gQG/MuAGyt3yClnHTtfX6zg8/oCytMTESypYnfiH /HdtP+RQFO7yWNe8+K1FUawe3RDwicHElDcRkAryP8OnHX+t7IjOaw3l1A6M9i7eL/VoPC1xQEB ynf8N94pvAXzY6QGhPKfN/GB2d3p10IoPcrCNPReNtweV44a06ZA3NSYe91tA4a4Nv9yQz6Jl5a NalnYdwsT6tke5W3PmCFB4nK4Mr4PT X-Gm-Gg: ASbGncsncG6sDWgUSH7SbMFjbBjVbjOL0hDDt9brcDxeDtydE5pg+anfXTsrVP7RdtJ 1MN8nDm8c1KTEvydzZtHAk0I1iiUijg83AG25mhiW0bssog1oZqjMD55HNol/nzHD0FORHjDrt6 tgbA3SHSAxCPzbjCIjOfut37V17kb2mP+5YCxSh1a61veYB4rhRyTm7nVb X-Received: by 2002:a05:6402:50cd:b0:63e:2d4d:5a60 with SMTP id 4fb4d7f45d1cf-64061a2067cmr2022068a12.8.1761812765799; Thu, 30 Oct 2025 01:26:05 -0700 (PDT) X-Google-Smtp-Source: AGHT+IEavuTsYWwslRUOAOxE/zFIdskj8vpMpsen2sBGmSlW7lcV0KGxaTvPZ688eyTXOVj7Oh7NHK0D25Uz4+668PI= X-Received: by 2002:a05:6402:50cd:b0:63e:2d4d:5a60 with SMTP id 4fb4d7f45d1cf-64061a2067cmr2022041a12.8.1761812765050; Thu, 30 Oct 2025 01:26:05 -0700 (PDT) MIME-Version: 1.0 References: <20251017062838.21041-1-yuhuang@redhat.com> <20251017062838.21041-5-yuhuang@redhat.com> <20251024010438.16e19757@elisabeth> <20251028124426.534fd236@elisabeth> <20251029053804.711716aa@elisabeth> <20251029080953.4f671f56@elisabeth> <20251029083929.0dc52bc6@elisabeth> <20251029131808.3e2a70c9@elisabeth> In-Reply-To: <20251029131808.3e2a70c9@elisabeth> From: Yumei Huang Date: Thu, 30 Oct 2025 16:25:53 +0800 X-Gm-Features: AWmQ_bka0ob4gXHB0EugjnIW1cJ3XGofguBUQSdTBdFEUrQ7WlzH5TKHtD4yWaI Message-ID: Subject: Re: [PATCH v6 4/4] tcp: Update data retransmission timeout To: Stefano Brivio X-Mimecast-Spam-Score: 0 X-Mimecast-MFC-PROC-ID: rD6xMlxrifbPnQEhjZbltGhfN0SERQz4aqAGPKzXvkk_1761812766 X-Mimecast-Originator: redhat.com Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Message-ID-Hash: 3JUPMSDFVKYVIFDAW2H2KOVT3WXCT5QK X-Message-ID-Hash: 3JUPMSDFVKYVIFDAW2H2KOVT3WXCT5QK X-MailFrom: yuhuang@redhat.com 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, david@gibson.dropbear.id.au 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: On Wed, Oct 29, 2025 at 8:18=E2=80=AFPM Stefano Brivio = wrote: > > On Wed, 29 Oct 2025 16:59:51 +0800 > Yumei Huang wrote: > > > On Wed, Oct 29, 2025 at 3:39=E2=80=AFPM Stefano Brivio wrote: > > > > > > On Wed, 29 Oct 2025 15:32:33 +0800 > > > Yumei Huang wrote: > > > > > > > On Wed, Oct 29, 2025 at 3:10=E2=80=AFPM Stefano Brivio wrote: > > > > > > > > > > On Wed, 29 Oct 2025 13:11:48 +0800 > > > > > Yumei Huang wrote: > > > > > > > > > > > On Wed, Oct 29, 2025 at 12:38=E2=80=AFPM Stefano Brivio wrote: > > > > > > > > > > > > > > On Wed, 29 Oct 2025 11:06:44 +0800 > > > > > > > Yumei Huang wrote: > > > > > > > > > > > > > > > On Tue, Oct 28, 2025 at 7:44=E2=80=AFPM Stefano Brivio wrote: > > > > > > > > > > > > > > > > > > On Tue, 28 Oct 2025 16:09:03 +0800 > > > > > > > > > Yumei Huang wrote: > > > > > > > > > > > > > > > > > > > On Fri, Oct 24, 2025 at 7:04=E2=80=AFAM Stefano Brivio = wrote: > > > > > > > > > > > > > > > > > > > > > > On Fri, 17 Oct 2025 14:28:38 +0800 > > > > > > > > > > > Yumei Huang wrote: > > > > > > > > > > > > > > > > > > > > > > > Use an exponential backoff timeout for data retrans= mission according > > > > > > > > > > > > to RFC 2988 and RFC 6298. Set the initial RTO to on= e second as discussed > > > > > > > > > > > > in Appendix A of RFC 6298. > > > > > > > > > > > > > > > > > > > > > > > > Also combine the macros defining the initial RTO fo= r both SYN and ACK. > > > > > > > > > > > > > > > > > > > > > > > > Signed-off-by: Yumei Huang > > > > > > > > > > > > Reviewed-by: David Gibson > > > > > > > > > > > > --- > > > > > > > > > > > > tcp.c | 27 ++++++++++++--------------- > > > > > > > > > > > > 1 file changed, 12 insertions(+), 15 deletions(-) > > > > > > > > > > > > > > > > > > > > > > > > diff --git a/tcp.c b/tcp.c > > > > > > > > > > > > index 9385132..dc0ec6c 100644 > > > > > > > > > > > > --- a/tcp.c > > > > > > > > > > > > +++ b/tcp.c > > > > > > > > > > > > @@ -179,16 +179,14 @@ > > > > > > > > > > > > * > > > > > > > > > > > > * Timeouts are implemented by means of timerfd ti= mers, set based on flags: > > > > > > > > > > > > * > > > > > > > > > > > > - * - SYN_TIMEOUT_INIT: if no ACK is received from = tap/guest during handshake > > > > > > > > > > > > - * (flag ACK_FROM_TAP_DUE without ESTABLISHED ev= ent) within this time, resend > > > > > > > > > > > > - * SYN. It's the starting timeout for the first = SYN retry. If this persists > > > > > > > > > > > > - * for more than TCP_MAX_RETRIES or (tcp_syn_ret= ries + > > > > > > > > > > > > - * tcp_syn_linear_timeouts) times in a row, rese= t the connection > > > > > > > > > > > > - * > > > > > > > > > > > > - * - ACK_TIMEOUT: if no ACK segment was received f= rom tap/guest, after sending > > > > > > > > > > > > - * data (flag ACK_FROM_TAP_DUE with ESTABLISHED = event), re-send data from the > > > > > > > > > > > > - * socket and reset sequence to what was acknowl= edged. If this persists for > > > > > > > > > > > > - * more than TCP_MAX_RETRIES times in a row, res= et the connection > > > > > > > > > > > > + * - RTO_INIT: if no ACK segment was received from= tap/guest, either during > > > > > > > > > > > > + * handshake (flag ACK_FROM_TAP_DUE without ESTA= BLISHED event) or after > > > > > > > > > > > > + * sending data (flag ACK_FROM_TAP_DUE with ESTA= BLISHED event), re-send data > > > > > > > > > > > > + * from the socket and reset sequence to what wa= s acknowledged. This is the > > > > > > > > > > > > + * timeout for the first retry, in seconds. If t= his persists too many times > > > > > > > > > > > > + * in a row, reset the connection: TCP_MAX_RETRI= ES for established > > > > > > > > > > > > + * connections, or (tcp_syn_retries + tcp_syn_li= near_timeouts) during the > > > > > > > > > > > > + * handshake. > > > > > > > > > > > > * > > > > > > > > > > > > * - FIN_TIMEOUT: if a FIN segment was sent to tap= /guest (flag ACK_FROM_TAP_DUE > > > > > > > > > > > > * with TAP_FIN_SENT event), and no ACK is recei= ved within this time, reset > > > > > > > > > > > > @@ -342,8 +340,7 @@ enum { > > > > > > > > > > > > #define WINDOW_DEFAULT 14600= /* RFC 6928 */ > > > > > > > > > > > > > > > > > > > > > > > > #define ACK_INTERVAL 10 = /* ms */ > > > > > > > > > > > > -#define SYN_TIMEOUT_INIT 1 = /* s */ > > > > > > > > > > > > -#define ACK_TIMEOUT 2 > > > > > > > > > > > > +#define RTO_INIT 1 = /* s, RFC 6298 */ > > > > > > > > > > > > #define FIN_TIMEOUT 60 > > > > > > > > > > > > #define ACT_TIMEOUT 7200 > > > > > > > > > > > > > > > > > > > > > > > > @@ -588,13 +585,13 @@ static void tcp_timer_ctl(con= st struct ctx *c, struct tcp_tap_conn *conn) > > > > > > > > > > > > } else if (conn->flags & ACK_FROM_TAP_DUE) { > > > > > > > > > > > > if (!(conn->events & ESTABLISHED)) { > > > > > > > > > > > > if (conn->retries < c->tcp.sy= n_linear_timeouts) > > > > > > > > > > > > - it.it_value.tv_sec = =3D SYN_TIMEOUT_INIT; > > > > > > > > > > > > + it.it_value.tv_sec = =3D RTO_INIT; > > > > > > > > > > > > else > > > > > > > > > > > > - it.it_value.tv_sec = =3D SYN_TIMEOUT_INIT << > > > > > > > > > > > > + it.it_value.tv_sec = =3D RTO_INIT << > > > > > > > > > > > > (conn->retrie= s - c->tcp.syn_linear_timeouts); > > > > > > > > > > > > } > > > > > > > > > > > > else > > > > > > > > > > > > - it.it_value.tv_sec =3D ACK_TI= MEOUT; > > > > > > > > > > > > + it.it_value.tv_sec =3D RTO_IN= IT << conn->retries; > > > > > > > > > > > > > > > > > > > > > > Same as on 3/4, but here it's clearly more convenient= : just assign > > > > > > > > > > > RTO_INIT, and multiply as needed in the if / else cla= uses. > > > > > > > > > > > > > > > > > > > > I guess we can't just assign RTO_INIT. Maybe assign it= only when > > > > > > > > > > retries=3D=3D0, otherwise multiply as it.it_value.tv_se= c <<=3D1. > > > > > > > > > > > > > > > > > > Why can't you do that? Say: > > > > > > > > > > > > > > > > > > it.it_value.tv_sec =3D RTO_INIT > > > > > > > > > if (!(conn->events & ESTABLISHED)) { > > > > > > > > > if (conn->retries >=3D c->tcp.syn_linear_= timeouts) > > > > > > > > > it.it_value.tv_sec <<=3D (conn->r= etries - > > > > > > > > > c->tcp.sy= n_linear_timeouts); > > > > > > > > > > > > > > > > > > but anyway, see below. > > > > > > > > > > > > > > > > > > > But it seems more complicated. What do you think? > > > > > > > > > > > > > > > > > > Or maybe, building on my latest comment to 3/4: > > > > > > > > > > > > > > > > > > int factor =3D conn->retries; > > > > > > > > > > > > > > > > > > if (!(conn->events & ESTABLISHED)) > > > > > > > > > factor -=3D c->tcp.syn_linear_tim= eouts; > > > > > > > > > > > > > > > > > > it.it_value.tv_sec =3D RTO_INIT << MAX(fa= ctor, 0); > > > > > > > > > > > > > > > > > > ? > > > > > > > > > > > > > > > > Yeah, I understand this part now. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > } else if (CONN_HAS(conn, SOCK_FIN_SENT | TAP= _FIN_ACKED)) { > > > > > > > > > > > > it.it_value.tv_sec =3D FIN_TIMEOUT; > > > > > > > > > > > > } else { > > > > > > > > > > > > > > > > > > > > > > The rest of the series looks good to me. > > > > > > > > > > > > > > > > > > > > > > It might be slightly more practical to factor in dire= ctly the RTO > > > > > > > > > > > clamp, and I don't think it's complicated now that yo= u have the helper > > > > > > > > > > > from 2/4, but it's not a strong preference from my si= de, as the series > > > > > > > > > > > makes sense in any case. > > > > > > > > > > > > > > > > > > > > Reading tcp_rto_max_ms can be easy with the helper. My = concern is > > > > > > > > > > about the way we get the total time for retries. > > > > > > > > > > > > > > > > > > > > I used to do it like this in v2, > > > > > > > > > > https://archives.passt.top/passt-dev/20251010074700.221= 77-4-yuhuang@redhat.com/: > > > > > > > > > > > > > > > > > > > > +#define RETRY_ELAPSED(timeout_init, retries) \ > > > > > > > > > > + ((timeout_init) * ((1 << ((retries) + 1)) - 2)) > > > > > > > > > > > > > > > > > > > > Though the formula is not quite right, we could refine = it as below: > > > > > > > > > > > > > > > > > > > > #define RETRY_ELAPSED(retries) ((RTO_INIT) * ((1 << ((r= etries) + 1)) - 1)) > > > > > > > > > > > > > > > > > > > > Does it make sense to get the time this way? > > > > > > > > > > > > > > > > > > Well, it also depends on c->tcp.syn_linear_timeouts, righ= t? > > > > > > > > > > > > > > > > Not really, it's only used for data retransmission, so > > > > > > > > syn_linear_timeouts is not relevant. > > > > > > > > > > > > > > Hmm, no, why? RFC 6298 covers SYN retries as well, and that's= the one > > > > > > > stating: > > > > > > > > > > > > I meant RETRY_ELAPSED was only used for data retransmission, wh= ich > > > > > > uses exponential backoff timeout directly, so syn_linear_timeou= ts was > > > > > > not relevant. > > > > > > > > > > Ah, okay. > > > > > > > > > > > > > > > > > > > (2.5) A maximum value MAY be placed on RTO provided it is = at least 60 > > > > > > > seconds. > > > > > > > > > > > > For SYN retries, as we used linear backoff + exponential backof= f, and > > > > > > also limited by TCP_MAX_RETRIES, the possible max RTO is far le= ss than > > > > > > 60s. So we didn't clamp it. Do you think we need to clamp it as= well? > > > > > > > > > > If syn_linear_timeouts is 0 and tcp_syn_retries is 7, I guess we'= ll > > > > > reach 2^0 + 2^1 + 2^2 + 2^3 + 2^4 + 2^5 + 2^6 + 2^7 =3D 247 secon= ds? Or > > > > > just up to ... + 2^6, that is, 119 seconds? > > > > > > > > > > In any case, what is the difference compared to data retransmissi= ons? > > > > > > > > You are right. I assumed wrongly that syn_linear_timeouts is always= 4. > > > > When it's 0, it's the same as data retransmissions. > > > > > > > > > > Don't we have 3 bits to store the retry count as well, so we're l= imited > > > > > by TCP_MAX_RETRIES anyway? Looking at patch 1/4 I'd say it's the = same > > > > > counter. > > > > > > > > Yes, it's the same counter. I guess you mean we should clamp it as = well. > > > > > > I didn't check this part, I thought we already did, but if we don't, = we > > > should do that, yes. > > > > > > > > > > ...the only thing that I don't see implemented in this versio= n of the > > > > > > > patch is paragraph 5.7: > > > > > > > > > > > > > > (5.7) If the timer expires awaiting the ACK of a SYN segme= nt and the > > > > > > > TCP implementation is using an RTO less than 3 secon= ds, the RTO > > > > > > > MUST be re-initialized to 3 seconds when data transm= ission > > > > > > > begins (i.e., after the three-way handshake complete= s). > > > > > > > > > > > > > > I missed that while reviewing earlier versions. I guess we ne= ed to use > > > > > > > a MAX(x, 3) clamp if (c->conn->events & ESTABLISHED). I think= it's > > > > > > > simpler than re-introducing separate starting values (one sec= ond and > > > > > > > three seconds). > > > > Sorry I'd like to confirm one more thing. By "use MAX(x, 3) clamp if > > (c->conn->events & ESTABLISHED)", I guess you mean: > > > > it.it_value.tv_sec =3D MAX(it.it_value.tv_sec, 3); > > Right. > > > Then the code would be: > > > > it.it_value.tv_sec =3D RTO_INIT; > > if (!(conn->events & ESTABLISHED)) { > > int exp =3D conn->retries - c->tcp.syn_linear_timeouts; > > it.it_value.tv_sec <<=3D MAX(exp, 0); > > } else { > > it.it_value.tv_sec =3D MAX(it.it_value.tv_sec, 3); > > it.it_value.tv_sec <<=3D conn->retries; > > Hmm what's wrong with my previous suggestion of initialising 'exp' with > conn->retries? Then it becomes (equivalent to your code snippet, but > shorter): Sorry, I read the comments several times but still missed that. And yes, it's shorter. > > > } > > it.it_value.tv_sec =3D MIN( > > it.it_value.tv_sec, c->tcp.tcp_rto_max); > > something like: > > int exp =3D conn->retries, min =3D 0, max =3D c->tcp.tcp_= rto_max; > > if (!(conn->events & ESTABLISHED)) > exp -=3D c->tcp.syn_linear_timeouts; > else > min =3D /* add a constant for this */ 3; > > it.it_value.tv_sec =3D RTO_INIT << MAX(exp, 0); > > it.it_value.tv_sec =3D CLAMP(it.it_value.tv_sec, min, max= ); But this seems not right. For data retransmission, only the first timeout will be changed from 1 to 3s, then the timeout becomes 4, 8, 16 ..., 120. I feel we don't need CLAMP, the code would be: #define RTO_INIT_ACK 3 ... int exp =3D conn->retries; it.it_value.tv_sec =3D RTO_INIT; if (!(conn->events & ESTABLISHED)) exp -=3D c->tcp.syn_linear_timeouts; else { it.it_value.tv_sec =3D MAX(it.it_value.tv_sec, RTO_INIT_ACK); } it.it_value.tv_sec <<=3D MAX(exp, 0); it.it_value.tv_sec =3D MIN( it.it_value.tv_sec, c->tcp.tcp_rto_max); Please correct me if I got anything wrong. Thanks. > > ...we don't have a CLAMP() macro at the moment, just MIN() and MAX() in > util.h, but now that we need one, I would add it, similar to the one > from the Linux kernel but without type checking (as it's not really > practical here). > > Inspired from include/linux/kernel.h (current Linux kernel tree): > > #define clamp(val, lo, hi) min((typeof(val))max(val, lo), hi) > > we can do something similar with two differences: 1. all our macros are > uppercase (we don't have as many as the kernel, and it's nice to know > that they are macros from the name) and 2. as I mentioned we don't > need/want type checking, so, say: > > #define CLAMP(x, min, max) MIN(MAX((x), (min)), (max)) > > ...or keep val / lo / hi if it's clearer, I don't really have a > preference. > > I added parentheses around the arguments by the way because I think > it's good practice, even though not needed in this case. It's the > PRE02-C recommendation in CERT C (we generally stick to those > recommendations whenever practical): > > https://wiki.sei.cmu.edu/confluence/display/c/PRE02-C.+Macro+replacemen= t+lists+should+be+parenthesized > > https://clang.llvm.org/extra/clang-tidy/checks/bugprone/macro-parenthes= es.html > > Note that the Linux kernel has a compatible license (it's the same, > actually, GPLv2+), and, regardless of that, the implementation is > trivial and the idea is obvious, so I don't think we need to give > explicit attribution in this case. > > But in other cases we do, or I guess it's fair to the author anyway, > see for example siphash.h. > > > we basically set the starting value to 3 for data retransmissions no > > matter if we have retried SYN. And the max total timeout would be > > 3+6+12+24+48+96+120+120 =3D 429s. Is it what we want? > > Ah, yes. In the discussion I previously assumed that the default clamp > value was 60 seconds, but it's actually 120 seconds. It looks correct > to me. > > > Besides, I guess I > > need to define a macro for "3" as well, like "ACK_TIMEOUT_INIT"? > > Or RTO_INIT_ACK? Maybe RTO_INIT_DATA? I'm thinking that you already > have a RTO_INIT at this point, and this constant is very closely related > to that one. > > > > > > > > > > > > > I'm not sure I understand here. If the timer expires, didn't we= reset > > > > > > the connection directly? We would never get to the data transmi= ssion > > > > > > phase? > > > > > > > > > > In the RFC, "the timer expires" indicates one iteration of the ti= meout > > > > > algorithm, that is, it's the same as our timer (tcp_timer_handler= ()) > > > > > triggering. > > > > > > > > I see. > > > > > > > > If you agree, I can add a new patch in this series to address the c= lamping. > > > > > > I don't really have a preference, I guess it could be directly in 4/4= or > > > in a separate patch, neither option should complicate things too much= . > > > > > > > > The paragraph says "after the three-way handshake completes", so = it > > > > > looks like the connection wasn't reset. > > > > > > > > > > > > > Probably I should name it more clearly. > > > > > > > > > > > > > > > > > > > > > > > > > > But in any case, do you really need to calculate that exp= licitly? I was > > > > > > > > > thinking that you can just clamp the value when you use i= t in > > > > > > > > > tcp_timer_ctl(). > > > > > > > > > > > > > > > > > > If c->tcp.rto_max is DIV_ROUND_CLOSEST(x, 1000), where 'x= ' is the value > > > > > > > > > you read from the kernel, then I guess it's just: > > > > > > > > > > > > > > > > > > --- > > > > > > > > > it.it_value.tv_sec =3D MIN(it.it_value.tv= _sec, c->tcp.rto_max); > > > > > > > > > > > > > > > > After reading the comments in v3 when tcp_rto_max_ms was fi= rst > > > > > > > > mentioned again, I realized I got something wrong again. I = thought it > > > > > > > > was for the total timeout for all retries, so I need to cal= culate that > > > > > > > > and decide to reset the connection as in v2. > > > > > > > > > > > > > > I think it actually applies to all the retries. > > > > > > > > > > > > > > > Anyway, you are right. We don't need to do that. Thanks for= your patience. > > > > > > > > > > > > > > > > > } ... > > > > > > > > > > > > > > > > > > if (timerfd_settime(conn->timer, 0, &it, NULL)) > > > > > > > > > flow_perror(conn, "failed to set timer"); > > > > > > > > > --- > > -- > Stefano > --=20 Thanks, Yumei Huang