[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <65411680-edf9-4670-be49-b73cc845c1cf@auristor.com>
Date: Fri, 10 Nov 2023 11:12:13 -0500
From: Jeffrey E Altman <jaltman@...istor.com>
To: David Howells <dhowells@...hat.com>
Cc: Marc Dionne <marc.dionne@...istor.com>,
linux-afs@...ts.infradead.org, linux-fsdevel@...r.kernel.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH 01/41] rxrpc: Fix RTT determination to use PING ACKs as a
source
On 11/10/2023 9:15 AM, Jeffrey E Altman wrote:
>
>> I'm not sure how expanding it internally to 64-bits actually helps
>> since the
>> upper 32 bits is not visible to the peer.
> rxrpc_complete_rtt_probe contains the following logic that relies on
> after() being able to detect
> a serial number wrap.
>
> /* If a later serial is being acked, then mark this slot as
> * being available.
> */
>
> if (after(acked_serial, orig_serial)) {
> trace_rxrpc_rtt_rx(call, rxrpc_rtt_rx_obsolete, i,
> orig_serial, acked_serial, 0, 0);
> clear_bit(i + RXRPC_CALL_RTT_PEND_SHIFT, &call->rtt_avail);
> smp_wmb();
> set_bit(i, &call->rtt_avail);
> }
>
> Otherwise, acked_serial = 0x01 will be considered smaller than
> orig_serial = 0xfffffffe and the slot will not be marked available.
>
> I will note that there is a similar problem with rxrpc_seq_t values
> which are u32 on the wire but which will wrap for calls that transmit
> more than approximately 5.5TB of data. Calls of this size are unlikely
> for a cache manager but are common for any service transmitting volume
> dumps.
I misread the definition of after() which is
static inline bool after(u32 seq1, u32 seq2)
{
return (s32)(seq1 - seq2) > 0;
}
This is sufficient to detect the serial number and sequence number wrapping.
What it doesn't provide is a method for rxrpc tracing and /proc file to
report on the total number of packets that have been sent or received on
a call or connection which might or might not be important depending
upon the use case.
Jeffrey Altman
Download attachment "smime.p7s" of type "application/pkcs7-signature" (4039 bytes)
Powered by blists - more mailing lists