From owner-svn-src-all@freebsd.org Fri Jun 21 00:24:53 2019 Return-Path: Delivered-To: svn-src-all@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2610:1c1:1:606c::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id D6F1215CBED0; Fri, 21 Jun 2019 00:24:52 +0000 (UTC) (envelope-from asomers@gmail.com) Received: from mail-lj1-f194.google.com (mail-lj1-f194.google.com [209.85.208.194]) (using TLSv1.3 with cipher TLS_AES_128_GCM_SHA256 (128/128 bits) server-signature RSA-PSS (4096 bits) client-signature RSA-PSS (2048 bits) client-digest SHA256) (Client CN "smtp.gmail.com", Issuer "GTS CA 1O1" (verified OK)) by mx1.freebsd.org (Postfix) with ESMTPS id 320496E67F; Fri, 21 Jun 2019 00:24:52 +0000 (UTC) (envelope-from asomers@gmail.com) Received: by mail-lj1-f194.google.com with SMTP id h10so4387477ljg.0; Thu, 20 Jun 2019 17:24:52 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=VlIboitvJ2toxKWj+darijoWRULHSEfG0NPmFZFetJc=; b=uDbevFyTES/T8eEOvpIRPWpB92IklOJZOp3zDcUfEUsPdNZdLS26YkSKtALjFxcseT 7g2g/uu/j+W3D0e+FoX2YMEt5vzSGsEJYo64/v26m1ek8g42HipWZRm7Fa1PGgUFGDts BBiAsu0ot+QRtHySrMHu5CsBwD9bK/8MS5C9G8JV1SlVQRdvyx02OFG6LKKu9smAJoxR ojUXWMCVFAAY9EaCy9Scws9UxLpl89R3/I8qiiz6V0wTDDUTAaq2KhhoWaytCjaDf+pi N9Ea5M62Xf7Po1h3f8spdvJMpY+jsDSngkyfn2s3E6MCRMTAAG5XPv4fBW6iA7KRNCBO 0PAg== X-Gm-Message-State: APjAAAX/G6I3qjB0hKbJOcowsCvkiwGA5dSJP/2mJg6zBJP6JOEctUFc KrEx/9i+Ud8ljDzXW+GqX+NYFvE2vypljuiET4H/kT3KzJ4= X-Google-Smtp-Source: APXvYqxNX3zCdeuQvVriihaw4PUhW7kCHZ7uBAidXhm5F1wK2CDw7dghuYSlbxbPIr8ctg2LVMZVArru0xODchFY8go= X-Received: by 2002:a2e:2c07:: with SMTP id s7mr31336989ljs.44.1561076690316; Thu, 20 Jun 2019 17:24:50 -0700 (PDT) MIME-Version: 1.0 References: <201906201435.x5KEZTqH021513@repo.freebsd.org> <54f3bc97cbb485cdcc44b81c82c149ac9e46d42f.camel@freebsd.org> <20190621013236.N5105@besplex.bde.org> <20190621033049.D5823@besplex.bde.org> In-Reply-To: From: Alan Somers Date: Thu, 20 Jun 2019 18:24:38 -0600 Message-ID: Subject: Re: svn commit: r349233 - head/sys/sys To: Warner Losh Cc: Bruce Evans , Ian Lepore , src-committers , svn-src-all , svn-src-head Content-Type: text/plain; charset="UTF-8" X-Rspamd-Queue-Id: 320496E67F X-Spamd-Bar: ------ Authentication-Results: mx1.freebsd.org X-Spamd-Result: default: False [-6.98 / 15.00]; REPLY(-4.00)[]; NEURAL_HAM_SHORT(-0.98)[-0.979,0]; NEURAL_HAM_MEDIUM(-1.00)[-0.997,0]; NEURAL_HAM_LONG(-1.00)[-1.000,0] X-BeenThere: svn-src-all@freebsd.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: "SVN commit messages for the entire src tree \(except for " user" and " projects" \)" List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 21 Jun 2019 00:24:53 -0000 On Thu, Jun 20, 2019 at 1:43 PM Warner Losh wrote: > > > > On Thu, Jun 20, 2019, 11:44 AM Bruce Evans wrote: >> >> On Thu, 20 Jun 2019, Alan Somers wrote: >> >> > On Thu, Jun 20, 2019 at 10:43 AM Bruce Evans wrote: >> >> Summary: and the headers that it includes should declare >> >> minimal types to compile (so __int64_t is enough). Most uses of this >> >> header require including domain-specific headers which declare the >> >> relevant data structures. >> > >> > Bruce, would you be satisfied by switching from to >> > and from int64_t to __int64_t? >> >> Not quite. The kernel block number type is daddr_t, and [__]int64_t is >> a hard coding of that. Hard-coding of typedefs is good for reducing >> namespace pollution, but it is not done for the nearby use of off_t. >> >> Unfortunately, daddr_t is only declared in . >> >> The type daddr_t seems to be correct, but ffs conflates it with >> ufs2_daddr_t. It is only for blocks of size DEV_BSIZE, while fs blocks >> are usually larger. The conflation is just a style bug because 64 bits >> should be large enough for anyone and ffs does the necessary conversions >> between DEV_BSIZE-blocks and fs-blocks. >> >> ext2fs seems to be more broken in this area. It doesn't have >> ext2fs_daddr_t, but hard-codes this as int32_t or int64_t. int32_t >> became very broken when ext2fs started attempting to support unsigned >> block numbers. Now, ext2_bmap() corrupts "daddr_t a_bn" to "int32_t >> bn" when it calls ext4_bmapext(). This has a chance of working via >> unsign-extension bugs provided a_bn fits in uint64_t. It is sure to >> fit for a valid a_bn, but perhaps your new ioctl allows probing for >> panics using invalid a_bn. ext2_bmap() also calls ext2_bmaparray(). >> That used to have essentially the same style bugs as ffs (with the >> ext2fs type corresponding to ufs2_daddr_t spelled int32_t), but it now >> correctly uses daddr_t. Internally, it misuses daddr_t for ext2fs >> block numbers, and there is an ext2fs block number type after all >> (e2fs_lbn_t = int64_t) which it uses only for metalbn. >> >> It looks like the older ext2fs code was fixed, especially if older ext2fs >> is limited to int32_t block numbers, but the ext4 case is quite broken >> since unsign extension bugs don't help it. a_bn starts as int64_t, then >> is truncated to the function arg "int32_t bn", then the function assigns bn >> to "int64_t lbn" and doesn't use bn again. >> >> Using a generic int64_t type in all interfaces would avoid some of these >> bugs, so I don't mind using it for the API. Just add a note that it must >> be large enough to represent all useful values of daddr_t. > > > Maybe we should add a __daddr_t define to sys/_types.h? And the usual reshuffling. That would also fix the namespace pollution. > > Warner See https://reviews.freebsd.org/D20715 .