From: Andreas Schwab Subject: Re: [PATCH 0/3] RISC-V SBI earlycon Date: Mon, 25 Mar 2019 17:23:15 +0100 Message-ID: References: <20181204135507.3706-1-anup@brainfault.org> Mime-Version: 1.0 Content-Type: text/plain Return-path: In-Reply-To: <20181204135507.3706-1-anup@brainfault.org> (Anup Patel's message of "Tue, 4 Dec 2018 19:25:04 +0530") Sender: linux-kernel-owner@vger.kernel.org To: Anup Patel Cc: Greg Kroah-Hartman , Jiri Slaby , Palmer Dabbelt , Albert Ou , Atish Patra , Christoph Hellwig , Rob Herring , linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, linux-serial@vger.kernel.org List-Id: linux-serial@vger.kernel.org On Dez 04 2018, Anup Patel wrote: > This patchset adds RISC-V SBI earlycon and removes RISC-V EARLY_PRINTK. > > We should use earlycon over existing EARLY_PRINTK for SBI console because: > 1. It's a more generic way of implementing early console for debugging > 2. Current RISC-V EARLY_PRINTK is a compile-time option whereas earlycon > is enabled at run-time via kernel parameters. > 3. To use earlycon with SBI, we have to pass "earlycon=sbi" in kernel > parameters. If earlycon kernel parameter is not provided then kernel > boots much faster which is very useful in real-world RISC-V deployments. Why doesn't this earlycon disable itself when the real console kicks in? Andreas. -- Andreas Schwab, SUSE Labs, schwab@suse.de GPG Key fingerprint = 0196 BAD8 1CE9 1970 F4BE 1748 E4D4 88E3 0EEA B9D7 "And now for something completely different."