September 2010 - kernel-russian@kernelnewbies.org

[PATCH] Introduce DMA xmit descriptor
by Denis Kirjanov Sept. 25, 2010

Sept. 25, 2010

Signed-off-by: Denis Kirjanov <kirjanov(a)gmail.com> --- >From datasheet TX/RX descriptors are very similar. I think that we can merge TX/RX in to one xmit descriptor. In this case we shoud use some kind of accessors to different fields. Any suggestions? Thanks! diff --git a/src/pcnet.c b/src/pcnet.c index 5a3e639..97acc7e 100644 --- a/src/pcnet.c +++ b/src/pcnet.c @@ -71,6 +71,20 @@ struct pcnet_dummy_init_block { __le32 tx_ring; } __attribute__ ((__packed__)); +/* 32bit TX/RX descriptors */ +/* PCnet Software Design Considerations, p.5 */ +struct xmit_descr { + __le32 addr; + /* The BCNT fields of the transmit and receive descriptors + * are 12-bit negative numbers representing the twos com- + * plement of the buffer size in bytes + */ + __le16 size; + __le16 status; + __le32 flags; + __le32 reserved; +}; + struct pcnet_private { /* TODO: * DMA buffer management operations

2 1

(no subject)
by Denis Kirjanov Sept. 24, 2010

Sept. 24, 2010

Signed-off-by: Denis Kirjanov <kirjanov(a)gmail.com> --- >From datasheet TX/RX descriptors are very similar. I think that we can merge TX/RX in to one xmit descriptor. In this case we shoud use some kind of accessors to different fields. Any suggestions? Thanks! diff --git a/src/pcnet.c b/src/pcnet.c index 5a3e639..97acc7e 100644 --- a/src/pcnet.c +++ b/src/pcnet.c @@ -71,6 +71,20 @@ struct pcnet_dummy_init_block { __le32 tx_ring; } __attribute__ ((__packed__)); +/* 32bit TX/RX descriptors */ +/* PCnet Software Design Considerations, p.5 */ +struct xmit_descr { + __le32 addr; + /* The BCNT fields of the transmit and receive descriptors + * are 12-bit negative numbers representing the twos com- + * plement of the buffer size in bytes + */ + __le16 size; + __le16 status; + __le32 flags; + __le32 reserved; +}; + struct pcnet_private { /* TODO: * DMA buffer management operations

1 0

[PATCH] Introduce initialization block
by Denis Kirjanov Sept. 23, 2010

Sept. 23, 2010

Introduce initialization block. Signed-off-by: Denis Kirjanov <kirjanov(a)gmail.com> --- src/pcnet.c | 43 +++++++++++++++++++++++++++++++++++++++++++ 1 files changed, 43 insertions(+), 0 deletions(-) diff --git a/src/pcnet.c b/src/pcnet.c index 537ca16..5a3e639 100644 --- a/src/pcnet.c +++ b/src/pcnet.c @@ -12,6 +12,7 @@ #include <linux/netdevice.h> #include <linux/etherdevice.h> #include <linux/spinlock.h> +#include <linux/types.h> #define DRV_NAME "pcnet_dummy" #define DRV_VERSION "dev" @@ -32,6 +33,44 @@ enum { PCNET_IOSIZE_LEN = 0x20, }; +/* PCnet-PCI II controller initialization includes the reading + * of the initialization block in memory to obtain the operat- + * ing parameters. + * When SSIZE32 (BCR20, bit 8) is set to ONE, all + * initialization block entries are logically 32-bits wide. + * + * The PCnet-PCI II controller obtains the start address of + * the initialization block from the contents of CSR1 (least + * significant 16 bits of address) and CSR2 (most signifi- + * cant 16 bits of address). The host must write CSR1 and + * CSR2 before setting the INIT bit. The initialization block + * contains the user defined conditions for PCnet-PCI II + * controller operation, together with the base addresses + * and length information of the transmit and receive + * descriptor rings. + */ + +/* Am79C970A reference, p. 69 */ +struct pcnet_dummy_init_block { + __le16 mode; + /* num of entries in TX/RX rings */ + __le16 txlen_rxlen; + /* MAC address */ + u8 mac_addr[6]; + __le16 reserved; + /* The Logical Address Filter (LADRF) is a programmable + * 64-bit mask that is used to accept incoming packets + * based on Logical (Multicast) Addresses. + */ + __le32 laddr_filter_low; + __le32 laddr_filter_hi; + /* programmed with the physical address of the receive + * and transmit descriptor rings + */ + __le32 rx_ring; + __le32 tx_ring; +} __attribute__ ((__packed__)); + struct pcnet_private { /* TODO: * DMA buffer management operations @@ -81,6 +120,10 @@ static int __devinit pcnet_dummy_init_netdev(struct pci_dev *pdev, /* init DMA rings */ /* init net_dev_ops */ + /* FIXME: + * isolate explicit HAVE_NET_DEVICE_OPS + * with a macro + */ #ifdef HAVE_NET_DEVICE_OPS ndev->netdev_ops = &pcnet_net_device_ops; #else -- 1.7.0.4

3 2

Драйвер сетевой карты
by Alex Naumov Sept. 22, 2010

Sept. 22, 2010

Всем привет. Есть идея написать простейший драйвер для сетевой карты. Если кого интересует, присоединяйтесь к обсуждению ;) Пока оттолкнуться можно от http://linuxgazette.net/156/jangir.html, http://linuxportal.ru/entry.php/2622_0_3_0_C/ и http://dmilvdv.narod.ru/Translate/index.html - Linux Device Drivers, Third Edition (LDD3) на русском языке С уважением, Alex

4 18

[PATCH] Continue pcnet probing. Introduce driver's private block
by Denis Kirjanov Sept. 18, 2010

Sept. 18, 2010

Signed-off-by: Denis Kirjanov <kirjanov(a)gmail.com> --- src/pcnet.c | 87 ++++++++++++++++++++++++++++++++++++++++++++++++---------- 1 files changed, 72 insertions(+), 15 deletions(-) diff --git a/src/pcnet.c b/src/pcnet.c index bdb463c..93fd4d1 100644 --- a/src/pcnet.c +++ b/src/pcnet.c @@ -9,6 +9,9 @@ #include <linux/init.h> #include <linux/kernel.h> #include <linux/pci.h> +#include <linux/netdevice.h> +#include <linux/etherdevice.h> +#include <linux/spinlock.h> #define DRV_NAME "pcnet_dummy" #define DRV_VERSION "dev" @@ -24,38 +27,86 @@ static DEFINE_PCI_DEVICE_TABLE(pcnet_dummy_pci_tbl) = { { } }; -static int pcnet_dummy_init_netdev(struct pci_dev *pdev, unsigned int ioaddr); +enum { + PCNET_IOSIZE_LEN = 0x20, +}; + +struct pcnet_private { + /* TODO: + * DMA buffer management operations + */ + spinlock_t lock; + struct pci_dev *pci_dev; + void __iomem *base; +}; + +/* Why do we need to create protos like this? */ +static int pcnet_dummy_init_netdev(struct pci_dev *pdev, unsigned long ioaddr); static int __devinit pcnet_dummy_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) { - unsigned int ioaddr; - int err; + struct pcnet_private *pp; + struct net_device *ndev; + void __iomem *ioaddr; +#ifdef USE_IO_OPS + int bar = 0; +#else + int bar = 1; +#endif if (pci_enable_device(pdev)) return -ENODEV; /* enables bus-mastering for device pdev */ pci_set_master(pdev); - if (pci_request_regions(pdev, DRV_NAME)) { - pci_disable_device(pdev); - return -EBUSY; + ndev = alloc_etherdev(sizeof(*pp)); + if (!ndev) { + goto out; } + /* register ourself under /sys/class/net/ */ + SET_NETDEV_DEV(ndev, &pdev->dev); - ioaddr = pci_resource_start(pdev, 0); + if (pci_request_regions(pdev, DRV_NAME)) + goto out_netdev; - err = pcnet_dummy_init_netdev(pdev, ioaddr); - if (err) { - pci_disable_device(pdev); - pci_release_regions(pdev); - } + ioaddr = pci_iomap(pdev, bar, PCNET_IOSIZE_LEN); + if (!ioaddr) + goto out_res; + pci_set_drvdata(pdev, ndev); + + if (pcnet_dummy_init_netdev(pdev, (unsigned long)ioaddr)) + goto out_res_unmap; - return err; + return 0; + +out_res_unmap: + pci_iounmap(pdev, ioaddr); +out_res: + pci_release_regions(pdev); +out_netdev: + free_netdev(ndev); +out: + pci_disable_device(pdev); + return -ENODEV; } static int __devinit pcnet_dummy_init_netdev(struct pci_dev *pdev, - unsigned int ioaddr) + unsigned long ioaddr) { + struct net_device *ndev = pci_get_drvdata(pdev); + struct pcnet_private *pp; + int irq; + + irq = pdev->irq; + pp = netdev_priv(ndev); + ndev->base_addr = ioaddr; + ndev->irq = irq; + pp->pci_dev = pdev; + spin_lock_init(&pp->lock); + + /* init DMA rings */ + /* init net_dev_ops */ /* registers net_device and returns err */ /* TODO: reset the chip at the end */ @@ -64,10 +115,16 @@ static int __devinit pcnet_dummy_init_netdev(struct pci_dev *pdev, static void __devexit pcnet_dummy_remove_one(struct pci_dev *pdev) { - /* TODO: ensure pdev is enabled */ + struct net_device *ndev = pci_get_drvdata(pdev); + struct pcnet_private *pp; + pp = netdev_priv(ndev); + pci_iounmap(pdev, pp->base); + free_netdev(ndev); + /* TODO: ensure pdev is enabled */ pci_disable_device(pdev); pci_release_regions(pdev); + pci_set_drvdata(pdev, NULL); } static struct pci_driver pcnet_dummy_driver = { -- 1.7.0.4

2 2

[PATCH] gitignore: ignore editor-backup files and generated from sources in src/
by Sergey V. Sept. 16, 2010

Sept. 16, 2010

Signed-off-by: Sergey V. <sftp.mtuci(a)gmail.com> --- .gitignore | 2 ++ src/.gitignore | 10 ++++++++++ 2 files changed, 12 insertions(+), 0 deletions(-) create mode 100644 .gitignore create mode 100644 src/.gitignore diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..f54eb28 --- /dev/null +++ b/.gitignore @@ -0,0 +1,2 @@ +*~ +\#*# diff --git a/src/.gitignore b/src/.gitignore new file mode 100644 index 0000000..92b7fba --- /dev/null +++ b/src/.gitignore @@ -0,0 +1,10 @@ +.tmp_versions + +*.o +*.ko +*.mod.c +*.o.cmd +*.ko.cmd + +Module.symvers +modules.order -- 1.7.1

1 0

[PATCH] Added some PCI stubs
by Denis Kirjanov Sept. 15, 2010

Sept. 15, 2010

Signed-off-by: Denis Kirjanov <kirjanov(a)gmail.com> --- src/pcnet.c | 35 ++++++++++++++++++++++++++++++++++- 1 files changed, 34 insertions(+), 1 deletions(-) diff --git a/src/pcnet.c b/src/pcnet.c index 7946394..f320ffd 100644 --- a/src/pcnet.c +++ b/src/pcnet.c @@ -2,22 +2,55 @@ #include <linux/module.h> #include <linux/init.h> +#include <linux/kernel.h> +#include <linux/pci.h> MODULE_AUTHOR("Dmitry Podgorny <pasis.ua(a)gmail.com>"); MODULE_DESCRIPTION("PCNet-PCI II/III Ethernet controller driver"); MODULE_VERSION("dev"); MODULE_LICENSE("GPL"); +#define DRV_NAME "pcnet_dummy" + +static DEFINE_PCI_DEVICE_TABLE(pcnet_dummy_pci_tbl) = { + { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_LANCE) }, + { } +}; + +static int __devinit pcnet_dummy_init_one(struct pci_dev *pdev, + const struct pci_device_id *ent) +{ + /* basically pci-skeleton stuff */ + /* reset the chip at the end */ + return 0; +} + +static void __devexit pcnet_dummy_remove_one(struct pci_dev *pdev) +{ + /* noop */ +} + +static struct pci_driver pcnet_dummy_driver = { + .name = DRV_NAME, + .id_table = pcnet_dummy_pci_tbl, + .probe = pcnet_dummy_init_one, + .remove = __devexit_p(pcnet_dummy_remove_one), +#if 0 + /* FIXME: add PM hooks */ +#endif +}; + static int __init pcnet_init(void) { printk(KERN_INFO "pcnet: module has loaded"); - return 0; + return pci_register_driver(&pcnet_dummy_driver); } static void __exit pcnet_exit(void) { printk(KERN_INFO "pcnet: module was unloaded"); + pci_unregister_driver(&pcnet_dummy_driver); } module_init(pcnet_init); -- 1.7.0.4

3 5

лимит на размер сообщения
by Alex Naumov Sept. 13, 2010

Sept. 13, 2010

Всем привет. Размер сообщения в рассылке не может быть больше 40KB. Этого более чем достаточно для простых текстовых сообщений, однако могут возникнуть проблемы с прикрепляемыми файлмами. Пожалуйста, если не прикрепляйте много файлов. Лучше разбить большое сообщение на несколько маленьких. С уважением, Alex

1 1

Codepage problem
by Степан Кузьмин Sept. 13, 2010

Sept. 13, 2010

Спасибо за помощь, Sergey V. (sftp.mtuci(a)gmail.com) -- Кузьмин Степан Веб разработчик. ООО <<Аматей>> Россия, г.Екатеринбург. Kuzmin Stepan Web developer. <<Amatei>> Ltd. Russian Federation, Yekaterinburg. tel. +7 (343) 283-04-29 skype: KuzminStepan twitter: @KuzminStepan e-mail: kyzminst(a)gmail.com -------------------------------------------- Q: Why is this email three sentences or less? A: http://three.sentenc.es

1 0

Re: Kernel-russian Digest, Vol 1, Issue 6
by Степан Кузьмин Sept. 13, 2010

Sept. 13, 2010

Yes, I have got questions (???) instead russian symbols. 2010/9/13 <kernel-russian-request(a)kernelnewbies.org>: > Send Kernel-russian mailing list submissions to > kernel-russian(a)kernelnewbies.org > > To subscribe or unsubscribe via the World Wide Web, visit > http://lists.kernelnewbies.org/mailman/listinfo/kernel-russian > or, via email, send a message with subject or body 'help' to > kernel-russian-request(a)kernelnewbies.org > > You can reach the person managing the list at > kernel-russian-owner(a)kernelnewbies.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Kernel-russian digest..." > > > Today's Topics: > > 1. Re:Kernel-russian Digest, Vol 1, Issue 4 (Sergey V.) > 2. Re:Kernel-russian Digest, Vol 1, Issue 3 (Alexander Naumov) > 3. Re:??????? ??????? ????? (Alexander Naumov) > 4. Re:??????? ??????? ????? (Alex Naumov) > 5. ???? (Alex Naumov) > 6. Re:???? (Alex Naumov) > 7. Re:??????? ??????? ????? (Dmitry Podgorny) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Sun, 12 Sep 2010 21:16:31 +0400 > From: "Sergey V." <sftp.mtuci(a)gmail.com> > Subject: Re: Kernel-russian Digest, Vol 1, Issue 4 > To: kernel-russian(a)kernelnewbies.org > Message-ID: <201009122116.31924.sftp.mtuci(a)gmail.com> > Content-Type: Text/Plain; charset="koi8-r" > > On Sunday 12 of September 2010 20:56:48 ?????? ??????? wrote: >> It seems to me, there are some problems with charset in this mailing >> list. I`ve got ??? instead russian symbols. >> I`m using utf8 as a default charset. >> > > I had the same problem. You subscribed to digest's, which broke charset. > > Go to link > http://lists.kernelnewbies.org/mailman/options/kernel-russian/kyzminst@gmai… > and turn off 'Set Digest Mode' option. After this you will get single mails from > other memebers with with correct charset. > > > > ------------------------------ > > Message: 2 > Date: Sun, 12 Sep 2010 20:12:24 +0200 > From: Alexander Naumov <anaumov(a)suse.de> > Subject: Re: Kernel-russian Digest, Vol 1, Issue 3 > To: kernel-russian(a)kernelnewbies.org > Message-ID: <20100912201224.58blg5ydcwsc4g8c(a)imap.suse.de> > Content-Type: text/plain; charset=UTF-8; DelSp="Yes"; format="flowed" > > Am Sun 12 Sep 2010 06:01:54 PM CEST schrieb ?????? ??????? > <kyzminst(a)gmail.com>: > >> Is there utf-8 charset? >> > > Yes, do you have any problem? > > Regards, > -- > Alexander Naumov, KDE Developer > > > > > ------------------------------ > > Message: 3 > Date: Sun, 12 Sep 2010 20:37:29 +0200 > From: Alexander Naumov <anaumov(a)suse.de> > Subject: Re: ??????? ??????? ????? > To: kernel-russian(a)kernelnewbies.org > Message-ID: <20100912203729.xocmcnjlwkgoossk(a)imap.suse.de> > Content-Type: text/plain; charset=UTF-8; DelSp="Yes"; format="flowed" > > Am Sun 12 Sep 2010 07:08:49 PM CEST schrieb Dmitry Podgorny > <pasis.ua(a)gmail.com>: > >> ? ??? ???????, ??? ????? ?????? ?????? ??????? ? ????, ? ????????? ???? >> ?????? ? ????? ???????????? ???????? ??????, ????? ???????? >> ????????????? ?????? ???????????? ? ????????? >> > > ??, ? ?????, ??? ?????? ????? ? ????????? ???? ?????, > (?????????????? ????????????? ??????) > > -- > Alexander Naumov, KDE Developer > > > > ------------------------------ > > Message: 4 > Date: Sun, 12 Sep 2010 20:58:00 +0200 > From: Alex Naumov <posix.ru(a)googlemail.com> > Subject: Re: ??????? ??????? ????? > To: kernel-russian(a)kernelnewbies.org > Message-ID: > <AANLkTinyG6Uq7xinsnOGJ1_zo%3Dh%3D-3LHtVyQJ3vqY%2Bwf(a)mail.gmail.com> > Content-Type: text/plain; charset="utf-8" > >> >> ?? ??? ????? ?????? ???????? ????????? ? ???? ??? ???????? ??????? ?? >> ??? ??????????? ? ?????? >> > > ??????? ?? ??? ? ????????! ?, ??????, ???? ?? ????? ??????????? ??? ??? ? > ???? ;) > > ???????, ??? ???? ?? ??????? ???????????? ?????-???? ??????????? (??????? > ????????) ???? > http://code.google.com/intl/ru > ??? ???????? ???-?? ???-?? ?????? ?????????? > > ????? ? ??-?????? ???? ???????????? ?????? ? build.opensuse.org ;) >

1 0