September 2015 - U-Boot

[U-Boot] [PATCH] net/eth: fix a bug in on_ethaddr()
by Gong Qianyu 29 Oct '15

29 Oct '15

The loop should check all ethenet devices, not only the first device, to set each specified ethaddr, or it'll cause failure when we use other devices. Signed-off-by: Gong Qianyu <Qianyu.Gong(a)freescale.com> diff --git a/net/eth.c b/net/eth.c index d3ec8d6..e3ef80e 100644 --- a/net/eth.c +++ b/net/eth.c @@ -677,6 +677,7 @@ static int on_ethaddr(const char *name, const char *value, enum env_op op, memset(dev->enetaddr, 0, 6); } } + dev = dev->next; } while (dev != eth_devices); return 0; -- 2.1.0.27.g96db324

4 6

[U-Boot] [PATCH] net: change the env name to use const
by Josh Wu 29 Oct '15

29 Oct '15

As we don't modify the 'name' parameter, so change it to const. Signed-off-by: Josh Wu <josh.wu(a)atmel.com> --- include/net.h | 4 ++-- net/eth.c | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/include/net.h b/include/net.h index d09bec9..133fee4 100644 --- a/include/net.h +++ b/include/net.h @@ -229,8 +229,8 @@ void eth_set_current(void); /* set nterface to ethcur var */ int eth_get_dev_index(void); /* get the device index */ void eth_parse_enetaddr(const char *addr, uchar *enetaddr); -int eth_getenv_enetaddr(char *name, uchar *enetaddr); -int eth_setenv_enetaddr(char *name, const uchar *enetaddr); +int eth_getenv_enetaddr(const char *name, uchar *enetaddr); +int eth_setenv_enetaddr(const char *name, const uchar *enetaddr); /* * Get the hardware address for an ethernet interface . diff --git a/net/eth.c b/net/eth.c index d3ec8d6..8e146d2 100644 --- a/net/eth.c +++ b/net/eth.c @@ -31,13 +31,13 @@ void eth_parse_enetaddr(const char *addr, uchar *enetaddr) } } -int eth_getenv_enetaddr(char *name, uchar *enetaddr) +int eth_getenv_enetaddr(const char *name, uchar *enetaddr) { eth_parse_enetaddr(getenv(name), enetaddr); return is_valid_ethaddr(enetaddr); } -int eth_setenv_enetaddr(char *name, const uchar *enetaddr) +int eth_setenv_enetaddr(const char *name, const uchar *enetaddr) { char buf[20]; -- 1.9.1

3 4

[U-Boot] [PATCH] net: bootp fix vci string on SPL-Boot
by Hannes Schmelzer 29 Oct '15

29 Oct '15

If CONFIG_CMD_DHCP is enabled, the vci (vendor-class-identifier) string isn't inserted into the bootp-packet during SPL stage because the CONFIG_BOOTP_VCI_STRING instead CONFIG_SPL_NET_VCI_STRING We fix this with testing for CONFIG_SPL_BUILD and testing for existing CONFIG_SPL_NET_VCI_STRING. Signed-off-by: Hannes Schmelzer <oe5hpm(a)oevsv.at> --- net/bootp.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/net/bootp.c b/net/bootp.c index 43466af..b547786 100644 --- a/net/bootp.c +++ b/net/bootp.c @@ -498,7 +498,9 @@ static int dhcp_extended(u8 *e, int message_type, struct in_addr server_ip, } #endif -#ifdef CONFIG_BOOTP_VCI_STRING +#if defined(CONFIG_SPL_BUILD) && defined(CONFIG_SPL_NET_VCI_STRING) + put_vci(e, CONFIG_SPL_NET_VCI_STRING); +#elif defined(CONFIG_BOOTP_VCI_STRING) put_vci(e, CONFIG_BOOTP_VCI_STRING); #endif -- 1.9.1

5 6

[U-Boot] [PATCH] smsc95xx: Use zero length packets when RX fifo is empty
by Stefan Brüns 29 Oct '15

29 Oct '15

Using NAKs on empty RX fifo for bulk in transfers is the right choice for a interrupt driven model, but U-Boot uses polling and expects an immediate answer if there is no incoming packet. Using ZLP Bulk In Response (BIR) mode avoids unexpected timeouts in the host controller driver. As ZLP mode is reset default, there is no need to set it. Signed-off-by: Stefan Brüns <stefan.bruens(a)rwth-aachen.de> --- The current setting results in timeouts in the dwc2 hc driver, if an expected network response is lost. This is often hidden by incoming broadcast packets, as these trigger the completion of the bulk in transfer as well. drivers/usb/eth/smsc95xx.c | 16 ---------------- 1 file changed, 16 deletions(-) diff --git a/drivers/usb/eth/smsc95xx.c b/drivers/usb/eth/smsc95xx.c index 6e80184..ae8bc18 100644 --- a/drivers/usb/eth/smsc95xx.c +++ b/drivers/usb/eth/smsc95xx.c @@ -531,22 +531,6 @@ static int smsc95xx_init_common(struct usb_device *udev, struct ueth_data *dev, if (ret < 0) return ret; - ret = smsc95xx_read_reg(udev, HW_CFG, &read_buf); - if (ret < 0) - return ret; - debug("Read Value from HW_CFG : 0x%08x\n", read_buf); - - read_buf |= HW_CFG_BIR_; - ret = smsc95xx_write_reg(udev, HW_CFG, read_buf); - if (ret < 0) - return ret; - - ret = smsc95xx_read_reg(udev, HW_CFG, &read_buf); - if (ret < 0) - return ret; - debug("Read Value from HW_CFG after writing " - "HW_CFG_BIR_: 0x%08x\n", read_buf); - #ifdef TURBO_MODE if (dev->pusb_dev->speed == USB_SPEED_HIGH) { burst_cap = DEFAULT_HS_BURST_CAP_SIZE / HS_USB_PKT_SIZE; -- 2.1.4

3 4

[U-Boot] [PATCH 1/1] net: phy: micrel: add support for KSZ8021RNL & KSZ8031RNL
by slemieux.tyco＠gmail.com 29 Oct '15

29 Oct '15

From: Sylvain Lemieux <slemieux(a)tycoint.com> This patch adds support for Micrel KSZ8021RNL & KSZ8031RNL. Signed-off-by: Sylvain Lemieux <slemieux(a)tycoint.com> --- Note: I am sending the patch using <slemieux.tyco(a)gmail.com>, until I can submit patch using my regular e-mail <slemieux(a)tycoint.com>. drivers/net/phy/micrel.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/net/phy/micrel.c b/drivers/net/phy/micrel.c index 49f444a..cbec928 100644 --- a/drivers/net/phy/micrel.c +++ b/drivers/net/phy/micrel.c @@ -22,6 +22,16 @@ static struct phy_driver KSZ804_driver = { .shutdown = &genphy_shutdown, }; +static struct phy_driver KSZ8031_driver = { + .name = "Micrel KSZ8021/KSZ8031", + .uid = 0x221550, + .mask = 0xfffff0, + .features = PHY_BASIC_FEATURES, + .config = &genphy_config, + .startup = &genphy_startup, + .shutdown = &genphy_shutdown, +}; + static struct phy_driver KSZ8081_driver = { .name = "Micrel KSZ8081", .uid = 0x221560, @@ -282,6 +292,7 @@ static struct phy_driver ksz9031_driver = { int phy_micrel_init(void) { phy_register(&KSZ804_driver); + phy_register(&KSZ8031_driver); phy_register(&KSZ8081_driver); #ifdef CONFIG_PHY_MICREL_KSZ9021 phy_register(&ksz9021_driver); -- 1.7.12.4

2 2

[U-Boot] [PATCH v2] net: parse DHCP options from overloaded file/sname fields
by Stefan Brüns 29 Oct '15

29 Oct '15

If Option 52 in the vendor option field signals overloading of the file and/or sname fields, these field may contain additional options. Formatting of file/sname contained options is the same as in the vendor options field, but without the leading magic. Signed-off-by: Stefan Brüns <stefan.bruens(a)rwth-aachen.de> Acked-by: Joe Hershberger <joe.hershberger(a)ni.com> --- v2: added commit message, no code changes net/bootp.c | 42 ++++++++++++++++++++++++++++++++---------- 1 file changed, 32 insertions(+), 10 deletions(-) diff --git a/net/bootp.c b/net/bootp.c index 72a956d..afb2bb8 100644 --- a/net/bootp.c +++ b/net/bootp.c @@ -768,14 +768,12 @@ void bootp_request(void) } #if defined(CONFIG_CMD_DHCP) -static void dhcp_process_options(uchar *popt, struct bootp_hdr *bp) +static void dhcp_process_options(uchar *popt, uchar *end) { - uchar *end = popt + BOOTP_HDR_SIZE; int oplen, size; #if defined(CONFIG_CMD_SNTP) && defined(CONFIG_BOOTP_TIMEOFFSET) int *to_ptr; #endif - dhcp_option_overload = 0; while (popt < end && *popt != 0xff) { oplen = *(popt + 1); @@ -860,6 +858,35 @@ static void dhcp_process_options(uchar *popt, struct bootp_hdr *bp) } } +static void dhcp_packet_process_options(struct bootp_hdr *bp) +{ + uchar *popt = (uchar *)&bp->bp_vend[4]; + uchar *end = popt + BOOTP_HDR_SIZE; + + if (net_read_u32((u32 *)&bp->bp_vend[0]) != htonl(BOOTP_VENDOR_MAGIC)) + return; + + dhcp_option_overload = 0; + + /* + * The 'options' field MUST be interpreted first, 'file' next, + * 'sname' last. + */ + dhcp_process_options(popt, end); + + if (dhcp_option_overload & OVERLOAD_FILE) { + popt = (uchar *)bp->bp_file; + end = popt + sizeof(bp->bp_file); + dhcp_process_options(popt, end); + } + + if (dhcp_option_overload & OVERLOAD_SNAME) { + popt = (uchar *)bp->bp_sname; + end = popt + sizeof(bp->bp_sname); + dhcp_process_options(popt, end); + } +} + static int dhcp_message_type(unsigned char *popt) { if (net_read_u32((u32 *)popt) != htonl(BOOTP_VENDOR_MAGIC)) @@ -977,14 +1004,11 @@ static void dhcp_handler(uchar *pkt, unsigned dest, struct in_addr sip, CONFIG_SYS_BOOTFILE_PREFIX, strlen(CONFIG_SYS_BOOTFILE_PREFIX)) == 0) { #endif /* CONFIG_SYS_BOOTFILE_PREFIX */ + dhcp_packet_process_options(bp); debug("TRANSITIONING TO REQUESTING STATE\n"); dhcp_state = REQUESTING; - if (net_read_u32((u32 *)&bp->bp_vend[0]) == - htonl(BOOTP_VENDOR_MAGIC)) - dhcp_process_options((u8 *)&bp->bp_vend[4], bp); - net_set_timeout_handler(5000, bootp_timeout_handler); dhcp_send_request_packet(bp); #ifdef CONFIG_SYS_BOOTFILE_PREFIX @@ -997,9 +1021,7 @@ static void dhcp_handler(uchar *pkt, unsigned dest, struct in_addr sip, debug("DHCP State: REQUESTING\n"); if (dhcp_message_type((u8 *)bp->bp_vend) == DHCP_ACK) { - if (net_read_u32((u32 *)&bp->bp_vend[0]) == - htonl(BOOTP_VENDOR_MAGIC)) - dhcp_process_options((u8 *)&bp->bp_vend[4], bp); + dhcp_packet_process_options(bp); /* Store net params from reply */ store_net_params(bp); dhcp_state = BOUND; -- 2.1.4

2 1

[U-Boot] [PATCH v2] net: Do not overwrite options found in overloaded 'file' field
by Stefan Brüns 29 Oct '15

29 Oct '15

If 'file' is overloaded, it is wrong to get or put the bootfile name from it/to it. Signed-off-by: Stefan Brüns <stefan.bruens(a)rwth-aachen.de> Acked-by: Joe Hershberger <joe.hershberger(a)ni.com> --- v2: added missing break before empty "case 53:", no functional change net/bootp.c | 41 +++++++++++++++-------------------------- 1 file changed, 15 insertions(+), 26 deletions(-) diff --git a/net/bootp.c b/net/bootp.c index 7fd29ee..72a956d 100644 --- a/net/bootp.c +++ b/net/bootp.c @@ -64,6 +64,9 @@ char net_root_path[64] = {0,}; /* Our bootpath */ static dhcp_state_t dhcp_state = INIT; static u32 dhcp_leasetime; static struct in_addr dhcp_server_ip; +static u8 dhcp_option_overload; +#define OVERLOAD_FILE 1 +#define OVERLOAD_SNAME 2 static void dhcp_handler(uchar *pkt, unsigned dest, struct in_addr sip, unsigned src, unsigned len); @@ -146,9 +149,11 @@ static void store_net_params(struct bootp_hdr *bp) net_copy_ip(&net_server_ip, &bp->bp_siaddr); memcpy(net_server_ethaddr, ((struct ethernet_hdr *)net_rx_packet)->et_src, 6); - if (strlen(bp->bp_file) > 0) + if (!(dhcp_option_overload & OVERLOAD_FILE) && + (strlen(bp->bp_file) > 0)) { copy_filename(net_boot_file_name, bp->bp_file, sizeof(net_boot_file_name)); + } debug("net_boot_file_name: %s\n", net_boot_file_name); @@ -770,6 +775,7 @@ static void dhcp_process_options(uchar *popt, struct bootp_hdr *bp) #if defined(CONFIG_CMD_SNTP) && defined(CONFIG_BOOTP_TIMEOFFSET) int *to_ptr; #endif + dhcp_option_overload = 0; while (popt < end && *popt != 0xff) { oplen = *(popt + 1); @@ -821,6 +827,9 @@ static void dhcp_process_options(uchar *popt, struct bootp_hdr *bp) case 51: net_copy_u32(&dhcp_leasetime, (u32 *)(popt + 2)); break; + case 52: + dhcp_option_overload = popt[2]; + break; case 53: /* Ignore Message Type Option */ break; case 54: @@ -832,31 +841,11 @@ static void dhcp_process_options(uchar *popt, struct bootp_hdr *bp) break; case 66: /* Ignore TFTP server name */ break; - case 67: /* vendor opt bootfile */ - /* - * I can't use dhcp_vendorex_proc here because I need - * to write into the bootp packet - even then I had to - * pass the bootp packet pointer into here as the - * second arg - */ - size = truncate_sz("Opt Boot File", - sizeof(bp->bp_file), - oplen); - if (bp->bp_file[0] == '\0' && size > 0) { - /* - * only use vendor boot file if we didn't - * receive a boot file in the main non-vendor - * part of the packet - god only knows why - * some vendors chose not to use this perfectly - * good spot to store the boot file (join on - * Tru64 Unix) it seems mind bogglingly crazy - * to me - */ - printf("*** WARNING: using vendor " - "optional boot file\n"); - memcpy(bp->bp_file, popt + 2, size); - bp->bp_file[size] = '\0'; - } + case 67: /* Bootfile option */ + size = truncate_sz("Bootfile", + sizeof(net_boot_file_name), oplen); + memcpy(&net_boot_file_name, popt + 2, size); + net_boot_file_name[size] = 0; break; default: #if defined(CONFIG_BOOTP_VENDOREX) -- 2.1.4

2 3

[U-Boot] [PATCH] net/arp: account for ARP delay, avoid duplicate packets on timeout
by Stefan Brüns 29 Oct '15

29 Oct '15

eth_rx() in the main reception loop may trigger sending a packet which is already timed out (or will immediately) upon reception of an ARP reply. As long as the ARP reply is pending, the timeout handler of a packet should be postponed. Happens on TFTP with bad network (e.g. WLAN). Signed-off-by: Stefan Brüns <stefan.bruens(a)rwth-aachen.de> --- net/arp.c | 5 +++-- net/arp.h | 2 +- net/net.c | 4 +++- 3 files changed, 7 insertions(+), 4 deletions(-) diff --git a/net/arp.c b/net/arp.c index b1f12bf..824d2e9 100644 --- a/net/arp.c +++ b/net/arp.c @@ -96,12 +96,12 @@ void arp_request(void) arp_raw_request(net_ip, net_null_ethaddr, net_arp_wait_reply_ip); } -void arp_timeout_check(void) +int arp_timeout_check(void) { ulong t; if (!net_arp_wait_packet_ip.s_addr) - return; + return 0; t = get_timer(0); @@ -118,6 +118,7 @@ void arp_timeout_check(void) arp_request(); } } + return 1; } void arp_receive(struct ethernet_hdr *et, struct ip_udp_hdr *ip, int len) diff --git a/net/arp.h b/net/arp.h index 43c6296..a288d61 100644 --- a/net/arp.h +++ b/net/arp.h @@ -25,7 +25,7 @@ void arp_init(void); void arp_request(void); void arp_raw_request(struct in_addr source_ip, const uchar *targetEther, struct in_addr target_ip); -void arp_timeout_check(void); +int arp_timeout_check(void); void arp_receive(struct ethernet_hdr *et, struct ip_udp_hdr *ip, int len); #endif /* __ARP_H__ */ diff --git a/net/net.c b/net/net.c index 2a6efd8..6c56e52 100644 --- a/net/net.c +++ b/net/net.c @@ -569,7 +569,9 @@ restart: goto done; } - arp_timeout_check(); + if (arp_timeout_check() > 0) { + time_start = get_timer(0); + } /* * Check for a timeout, and run the timeout handler -- 2.1.4

2 2

[U-Boot] [PATCH] net/arp: Do not run net_start_again() on timeout
by Stefan Brüns 29 Oct '15

29 Oct '15

net_start_again() will be called from net_loop() if state is NETLOOP_FAIL. Signed-off-by: Stefan Brüns <stefan.bruens(a)rwth-aachen.de> --- net/arp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/arp.c b/net/arp.c index b865570..b1f12bf 100644 --- a/net/arp.c +++ b/net/arp.c @@ -112,7 +112,7 @@ void arp_timeout_check(void) if (arp_wait_try >= ARP_TIMEOUT_COUNT) { puts("\nARP Retry count exceeded; starting again\n"); arp_wait_try = 0; - net_start_again(); + net_set_state(NETLOOP_FAIL); } else { arp_wait_timer_start = t; arp_request(); -- 2.1.4

2 2

[U-Boot] [PATCH] net: Fix parsing of Bootp/DHCP option 0 (Pad)
by Stefan Brüns 29 Oct '15

29 Oct '15

Pad has no len byte, so the normal parsing code fails. Signed-off-by: Stefan Brüns <stefan.bruens(a)rwth-aachen.de> --- net/bootp.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/net/bootp.c b/net/bootp.c index 9700e95..7fd29ee 100644 --- a/net/bootp.c +++ b/net/bootp.c @@ -774,6 +774,9 @@ static void dhcp_process_options(uchar *popt, struct bootp_hdr *bp) while (popt < end && *popt != 0xff) { oplen = *(popt + 1); switch (*popt) { + case 0: + oplen = -1; /* Pad omits len byte */ + break; case 1: net_copy_ip(&net_netmask, (popt + 2)); break; @@ -877,7 +880,13 @@ static int dhcp_message_type(unsigned char *popt) while (*popt != 0xff) { if (*popt == 53) /* DHCP Message Type */ return *(popt + 2); - popt += *(popt + 1) + 2; /* Scan through all options */ + if (*popt == 0) { + /* Pad */ + popt += 1; + } else { + /* Scan through all options */ + popt += *(popt + 1) + 2; + } } return -1; } -- 2.1.4

2 2