From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from mail-io0-x243.google.com ([2607:f8b0:4001:c06::243]) by bombadil.infradead.org with esmtps (Exim 4.90_1 #2 (Red Hat Linux)) id 1fJOgo-000331-1k for barebox@lists.infradead.org; Thu, 17 May 2018 19:35:51 +0000 Received: by mail-io0-x243.google.com with SMTP id f21-v6so3408845iob.13 for ; Thu, 17 May 2018 12:35:39 -0700 (PDT) MIME-Version: 1.0 In-Reply-To: <20180517094333.eyi6zjahrxd2vsqc@pengutronix.de> References: <20180516201827.6430-1-andrew.smirnov@gmail.com> <20180516201827.6430-3-andrew.smirnov@gmail.com> <20180517094333.eyi6zjahrxd2vsqc@pengutronix.de> From: Andrey Smirnov Date: Thu, 17 May 2018 12:35:38 -0700 Message-ID: List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: "barebox" Errors-To: barebox-bounces+u.kleine-koenig=pengutronix.de@lists.infradead.org Subject: Re: [PATCH v2 2/4] bootsource: Add bootsource alias name API To: Sascha Hauer Cc: Barebox List On Thu, May 17, 2018 at 2:43 AM, Sascha Hauer wrote: > On Wed, May 16, 2018 at 01:18:25PM -0700, Andrey Smirnov wrote: >> Add API allowing to query and override the name of the alias pointing >> at DTB node representing current bootsource. >> >> Signed-off-by: Andrey Smirnov >> --- >> common/bootsource.c | 74 ++++++++++++++++++++++++++++++++++++++++++++ >> include/bootsource.h | 2 ++ >> 2 files changed, 76 insertions(+) >> >> diff --git a/common/bootsource.c b/common/bootsource.c >> index 78ecd8267..ed435c051 100644 >> --- a/common/bootsource.c >> +++ b/common/bootsource.c >> @@ -41,6 +41,80 @@ static const char *bootsource_str[] = { >> >> static enum bootsource bootsource = BOOTSOURCE_UNKNOWN; >> static int bootsource_instance = BOOTSOURCE_INSTANCE_UNKNOWN; >> +const char *bootsource_alias_name = NULL; >> + >> +/** >> + * bootsource_get_alias_name() - Get the name of the bootsource alias >> + * >> + * This function will return newly allocated string containing name of >> + * the alias that is expected to point to DTB node corresponding to >> + * detected bootsource >> + * >> + * NOTE: Caller is expected to free() the string allocated by this >> + * function >> + */ >> +char *bootsource_get_alias_name(void) >> +{ >> + int instance = bootsource_instance; >> + const char *stem; >> + >> + /* >> + * If alias name was overridden via >> + * bootsource_set_alias_name() return that value without >> + * asking any questions. >> + * >> + * Note that we have to strdup() the result to make it >> + * free-able. >> + */ >> + if (bootsource_alias_name) >> + return strdup(bootsource_alias_name); >> + >> + if (bootsource >= ARRAY_SIZE(bootsource_str) || >> + bootsource == BOOTSOURCE_UNKNOWN) >> + return NULL; > > This is already handled by the default path of the switch/case below. > Good point. Will remove in v3. >> + >> + switch (bootsource) { >> + case BOOTSOURCE_I2C_EEPROM: /* FALLTHROUGH */ >> + case BOOTSOURCE_SPI_EEPROM: >> + /* >> + * For I2C and SPI EEPROMs we set the stem to be 'i2c' >> + * and 'spi' correspondingly. The resulting alias will >> + * be pointing at the controller said EEPROM is >> + * attached to. >> + * >> + * NOTE: This code assumes single bootable EEPROM per >> + * controller >> + */ >> + BUILD_BUG_ON(BOOTSOURCE_SPI != BOOTSOURCE_SPI_EEPROM - 1); >> + BUILD_BUG_ON(BOOTSOURCE_I2C != BOOTSOURCE_I2C_EEPROM - 1); >> + >> + stem = bootsource_str[bootsource - 1]; > > Can we make this conversion more explicit? > > case BOOTSOURCE_I2C_EEPROM: > stem = bootsource_str[BOOTSOURCE_I2C]; > break; > case BOOTSOURCE_SPI_EEPROM: > stem = bootsource_str[BOOTSOURCE_SPI]; > break; > > It makes it more obvious what happens here. > Sure, will do in v3. >> + break; >> + case BOOTSOURCE_SERIAL: /* FALLTHROUGH */ >> + case BOOTSOURCE_I2C: /* FALLTHROUGH */ >> + case BOOTSOURCE_MMC: /* FALLTHROUGH */ >> + case BOOTSOURCE_SPI: /* FALLTHROUGH */ >> + case BOOTSOURCE_CAN: >> + stem = bootsource_str[bootsource]; >> + break; >> + default: >> + return NULL; >> + } >> + >> + /* >> + * Assume 0 if instance was not specified by bootsource >> + * detection code. >> + */ >> + if (instance == BOOTSOURCE_INSTANCE_UNKNOWN) >> + instance = 0; > > Hm, if some SoC has forgotten to setup the instance I think we should > just bail out here and fix the SoC code instead. > OK, will do in v3. Thanks, Andrey Smirnov _______________________________________________ barebox mailing list barebox@lists.infradead.org http://lists.infradead.org/mailman/listinfo/barebox