From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 56A3F45D5F3 for ; Fri, 14 Aug 2026 10:19:30 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786702779; cv=none; b=dOussyoex40wGlEItBNPtkRijTy124NZRR9tTftaNJfZ45o7IpR+TJPBjQJ8vQYGX33gq0vVukIL7EQNuShYaGg+FO42lEJE4kIgGePpzqqjyP3xeVyIUuppZmkp9Roc38JmBuZ6OqbWWpXtApwbIGLHhkh//tTuDXu5DLzZ0/8= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786702779; c=relaxed/simple; bh=QIJ+zL/TA9DbcG66LAPLEYwrLzcjlPLrn7h0fCPyrLE=; h=From:Subject:To:Cc:In-Reply-To:References:Content-Type:Date: Message-Id; b=hY+qrwW0LknrgyEB/GJYZW5zApvWjY2tKZbVulXyvTq5pviVWMyYrqR+1lPE16qrRC8+58NtVGI+FEq6JzkGuTlQQkzE9yZnHXx1FUZRTxC7Q+Xw5+wrjaSpUQ14FKVp/ZfODdXPo0pdnMXuSZ6BC30qwNTPqjpxYf6Qiia2iQk= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=FyUGm0bi; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="FyUGm0bi" Received: by smtp.kernel.org (Postfix) with ESMTPSA id E94F51F000E9; Fri, 14 Aug 2026 10:19:27 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1786702768; bh=csxZ2Ig6IM/kB0D1UZzPL5LNWetuI5YiIKlZ+1dXrbQ=; h=From:Subject:Reply-To:To:Cc:In-Reply-To:References:Date; b=FyUGm0biFVV+rSVqsQ7R4dSO7N3fcrqhIpFvcZL19fwdjbMV8cT8SYpz2X++u5sl7 Sc70YvcxLVjlJksqWTMiV7ZfDzTDHGKQVFSnImH31dwMJ0lUgQBKbM4noJGdPpvDnC z9xuYw88yCZdhOXjzN6Ha+5zRUTkmc4nD15ZJMxppYlRffkn6MVSULvB0HbTpyOFpy 2N5px/b5VikSpnfZHmVsnjz9xOOXHGM8Ub9w2N4zLUTQ5b+wLMaJGtogV89EdmQ/RY A3H9N9RB1Zxzuygr1q4VS3yplw+wTpYcwGyF0QrsXeUKcTFXGaaa/cRPzrz7zEQMU2 GyOotWBl9UT5w== From: sashiko-bot@kernel.org Subject: Re: [PATCH -next v5 30/32] ext4: add tracepoints for EOF block zeroing and disksize-grow I/O Reply-To: sashiko-reviews@lists.linux.dev To: "Zhang Yi" Cc: linux-ext4@vger.kernel.org, tytso@mit.edu In-Reply-To: <20260814093331.1703882-31-yi.zhang@huaweicloud.com> References: <20260814093331.1703882-1-yi.zhang@huaweicloud.com> <20260814093331.1703882-31-yi.zhang@huaweicloud.com> Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Date: Fri, 14 Aug 2026 10:19:27 +0000 Message-Id: <20260814101927.E94F51F000E9@smtp.kernel.org> Precedence: bulk X-Mailing-List: linux-ext4@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Thank you for your contribution! Sashiko AI review found 1 potential issue(= s) to consider: - [High] Calling `i_size_read()` from the block I/O completion context (sof= tirq) inside the `ext4_iomap_wb_disksize_pending` tracepoint class causes a= hard lockup on 32-bit SMP systems. -- --- Patch [30]: [PATCH -next v5 30/32] ext4: add tracepoints for EOF block = zeroing and disksize-grow I/O --- commit 9f6116e7039af778ef1701a210ea9b800dcc543c Author: Zhang Yi ext4: add tracepoints for EOF block zeroing and disksize-grow I/O Add tracepoints to track the disksize-grow-pending lifecycle in the writeback path and the block-zero-EOF entry point. > diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c > index 4f1176b9332f2..4464eb03c9728 100644 > --- a/fs/ext4/page-io.c > +++ b/fs/ext4/page-io.c [ ... ] > @@ -729,8 +735,11 @@ void ext4_iomap_end_bio(struct bio *bio) > * state set in ext4_block_zero_eof() and wake up all waiters > * that will update the inode i_disksize. > */ > - if (io_mode =3D=3D EXT4_IOMAP_IOEND_DISKSIZE_GROW_IO) > + if (io_mode =3D=3D EXT4_IOMAP_IOEND_DISKSIZE_GROW_IO) { > + trace_ext4_iomap_wb_disksize_pending_complete(ioend->io_inode, > + ioend->io_offset, ioend->io_size); > ext4_iomap_clear_disksize_pending(ioend->io_inode); > + } > diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h > index 4539ef8e5f86c..c9259c2a3e364 100644 > --- a/include/trace/events/ext4.h > +++ b/include/trace/events/ext4.h [ ... ] > +/* disksize pending I/O tracepoints for iomap Buffered I/O path */ > +DECLARE_EVENT_CLASS(ext4_iomap_wb_disksize_pending, > + TP_PROTO(struct inode *inode, loff_t io_offset, size_t io_size), > + TP_ARGS(inode, io_offset, io_size), > + TP_STRUCT__entry( > + __field(dev_t, dev) > + __field(u64, ino) > + __field(loff_t, io_offset) > + __field(size_t, io_size) > + __field(loff_t, i_size) > + __field(loff_t, i_disksize) > + ), > + TP_fast_assign( > + __entry->dev =3D inode->i_sb->s_dev; > + __entry->ino =3D inode->i_ino; > + __entry->io_offset =3D io_offset; > + __entry->io_size =3D io_size; > + __entry->i_size =3D i_size_read(inode); [Severity: High] Does this code introduce a deadlock on 32-bit SMP systems? The tracepoint ext4_iomap_wb_disksize_pending_complete is called from ext4_iomap_end_bio(), which executes in softirq context. This happens because ext4_iomap_writeback_submit() clears the BIO_COMPLETE_IN_TASK flag. When TP_fast_assign() evaluates i_size_read(inode), it uses a read sequence lock on 32-bit SMP systems via read_seqcount_begin(&inode->i_size_seqcount). If i_size_write() is executing concurrently in process context on the same CPU, it holds the sequence lock without disabling softirqs. If the softirq interrupts the i_size_write() section, i_size_read() will spin infinitely waiting for the sequence count to become even. Is there a way to avoid reading the inode size in this softirq context, or a safer way to sample it here? > + __entry->i_disksize =3D READ_ONCE(EXT4_I(inode)->i_disksize); > + ), > + TP_printk("dev %d:%d ino %llu io_offset %lld io_size %zu i_size %lld i_= disksize %lld", > + MAJOR(__entry->dev), MINOR(__entry->dev), > + __entry->ino, __entry->io_offset, __entry->io_size, > + __entry->i_size, __entry->i_disksize) > +); --=20 Sashiko AI review =C2=B7 https://sashiko.dev/#/patchset/20260814093331.1703= 882-1-yi.zhang@huaweicloud.com?part=3D30