1=============================================================== 2HVCS IBM "Hypervisor Virtual Console Server" Installation Guide 3=============================================================== 4 5for Linux Kernel 2.6.4+ 6 7Copyright (C) 2004 IBM Corporation 8 9.. =========================================================================== 10.. NOTE:Eight space tabs are the optimum editor setting for reading this file. 11.. =========================================================================== 12 13 14Author(s): Ryan S. Arnold <rsa@us.ibm.com> 15 16Date Created: March, 02, 2004 17Last Changed: August, 24, 2004 18 19.. Table of contents: 20 21 1. Driver Introduction: 22 2. System Requirements 23 3. Build Options: 24 3.1 Built-in: 25 3.2 Module: 26 4. Installation: 27 5. Connection: 28 6. Disconnection: 29 7. Configuration: 30 8. Questions & Answers: 31 9. Reporting Bugs: 32 331. Driver Introduction: 34======================= 35 36This is the device driver for the IBM Hypervisor Virtual Console Server, 37"hvcs". The IBM hvcs provides a tty driver interface to allow Linux user 38space applications access to the system consoles of logically partitioned 39operating systems (Linux and AIX) running on the same partitioned Power5 40ppc64 system. Physical hardware consoles per partition are not practical 41on this hardware so system consoles are accessed by this driver using 42firmware interfaces to virtual terminal devices. 43 442. System Requirements: 45======================= 46 47This device driver was written using 2.6.4 Linux kernel APIs and will only 48build and run on kernels of this version or later. 49 50This driver was written to operate solely on IBM Power5 ppc64 hardware 51though some care was taken to abstract the architecture dependent firmware 52calls from the driver code. 53 54Sysfs must be mounted on the system so that the user can determine which 55major and minor numbers are associated with each vty-server. Directions 56for sysfs mounting are outside the scope of this document. 57 583. Build Options: 59================= 60 61The hvcs driver registers itself as a tty driver. The tty layer 62dynamically allocates a block of major and minor numbers in a quantity 63requested by the registering driver. The hvcs driver asks the tty layer 64for 64 of these major/minor numbers by default to use for hvcs device node 65entries. 66 67If the default number of device entries is adequate then this driver can be 68built into the kernel. If not, the default can be over-ridden by inserting 69the driver as a module with insmod parameters. 70 713.1 Built-in: 72------------- 73 74The following menuconfig example demonstrates selecting to build this 75driver into the kernel:: 76 77 Device Drivers ---> 78 Character devices ---> 79 <*> IBM Hypervisor Virtual Console Server Support 80 81Begin the kernel make process. 82 833.2 Module: 84----------- 85 86The following menuconfig example demonstrates selecting to build this 87driver as a kernel module:: 88 89 Device Drivers ---> 90 Character devices ---> 91 <M> IBM Hypervisor Virtual Console Server Support 92 93The make process will build the following kernel modules: 94 95 - hvcs.ko 96 - hvcserver.ko 97 98To insert the module with the default allocation execute the following 99commands in the order they appear:: 100 101 insmod hvcserver.ko 102 insmod hvcs.ko 103 104The hvcserver module contains architecture specific firmware calls and must 105be inserted first, otherwise the hvcs module will not find some of the 106symbols it expects. 107 108To override the default use an insmod parameter as follows (requesting 4 109tty devices as an example):: 110 111 insmod hvcs.ko hvcs_parm_num_devs=4 112 113There is a maximum number of dev entries that can be specified on insmod. 114We think that 1024 is currently a decent maximum number of server adapters 115to allow. This can always be changed by modifying the constant in the 116source file before building. 117 118NOTE: The length of time it takes to insmod the driver seems to be related 119to the number of tty interfaces the registering driver requests. 120 121In order to remove the driver module execute the following command:: 122 123 rmmod hvcs.ko 124 125The recommended method for installing hvcs as a module is to use depmod to 126build a current modules.dep file in /lib/modules/`uname -r` and then 127execute:: 128 129 modprobe hvcs hvcs_parm_num_devs=4 130 131The modules.dep file indicates that hvcserver.ko needs to be inserted 132before hvcs.ko and modprobe uses this file to smartly insert the modules in 133the proper order. 134 135The following modprobe command is used to remove hvcs and hvcserver in the 136proper order:: 137 138 modprobe -r hvcs 139 1404. Installation: 141================ 142 143The tty layer creates sysfs entries which contain the major and minor 144numbers allocated for the hvcs driver. The following snippet of "tree" 145output of the sysfs directory shows where these numbers are presented:: 146 147 sys/ 148 |-- *other sysfs base dirs* 149 | 150 |-- class 151 | |-- *other classes of devices* 152 | | 153 | `-- tty 154 | |-- *other tty devices* 155 | | 156 | |-- hvcs0 157 | | `-- dev 158 | |-- hvcs1 159 | | `-- dev 160 | |-- hvcs2 161 | | `-- dev 162 | |-- hvcs3 163 | | `-- dev 164 | | 165 | |-- *other tty devices* 166 | 167 |-- *other sysfs base dirs* 168 169For the above examples the following output is a result of cat'ing the 170"dev" entry in the hvcs directory:: 171 172 Pow5:/sys/class/tty/hvcs0/ # cat dev 173 254:0 174 175 Pow5:/sys/class/tty/hvcs1/ # cat dev 176 254:1 177 178 Pow5:/sys/class/tty/hvcs2/ # cat dev 179 254:2 180 181 Pow5:/sys/class/tty/hvcs3/ # cat dev 182 254:3 183 184The output from reading the "dev" attribute is the char device major and 185minor numbers that the tty layer has allocated for this driver's use. Most 186systems running hvcs will already have the device entries created or udev 187will do it automatically. 188 189Given the example output above, to manually create a /dev/hvcs* node entry 190mknod can be used as follows:: 191 192 mknod /dev/hvcs0 c 254 0 193 mknod /dev/hvcs1 c 254 1 194 mknod /dev/hvcs2 c 254 2 195 mknod /dev/hvcs3 c 254 3 196 197Using mknod to manually create the device entries makes these device nodes 198persistent. Once created they will exist prior to the driver insmod. 199 200Attempting to connect an application to /dev/hvcs* prior to insertion of 201the hvcs module will result in an error message similar to the following:: 202 203 "/dev/hvcs*: No such device". 204 205NOTE: Just because there is a device node present doesn't mean that there 206is a vty-server device configured for that node. 207 2085. Connection 209============= 210 211Since this driver controls devices that provide a tty interface a user can 212interact with the device node entries using any standard tty-interactive 213method (e.g. "cat", "dd", "echo"). The intent of this driver however, is 214to provide real time console interaction with a Linux partition's console, 215which requires the use of applications that provide bi-directional, 216interactive I/O with a tty device. 217 218Applications (e.g. "minicom" and "screen") that act as terminal emulators 219or perform terminal type control sequence conversion on the data being 220passed through them are NOT acceptable for providing interactive console 221I/O. These programs often emulate antiquated terminal types (vt100 and 222ANSI) and expect inbound data to take the form of one of these supported 223terminal types but they either do not convert, or do not _adequately_ 224convert, outbound data into the terminal type of the terminal which invoked 225them (though screen makes an attempt and can apparently be configured with 226much termcap wrestling.) 227 228For this reason kermit and cu are two of the recommended applications for 229interacting with a Linux console via an hvcs device. These programs simply 230act as a conduit for data transfer to and from the tty device. They do not 231require inbound data to take the form of a particular terminal type, nor do 232they cook outbound data to a particular terminal type. 233 234In order to ensure proper functioning of console applications one must make 235sure that once connected to a /dev/hvcs console that the console's $TERM 236env variable is set to the exact terminal type of the terminal emulator 237used to launch the interactive I/O application. If one is using xterm and 238kermit to connect to /dev/hvcs0 when the console prompt becomes available 239one should "export TERM=xterm" on the console. This tells ncurses 240applications that are invoked from the console that they should output 241control sequences that xterm can understand. 242 243As a precautionary measure an hvcs user should always "exit" from their 244session before disconnecting an application such as kermit from the device 245node. If this is not done, the next user to connect to the console will 246continue using the previous user's logged in session which includes 247using the $TERM variable that the previous user supplied. 248 249Hotplug add and remove of vty-server adapters affects which /dev/hvcs* node 250is used to connect to each vty-server adapter. In order to determine which 251vty-server adapter is associated with which /dev/hvcs* node a special sysfs 252attribute has been added to each vty-server sysfs entry. This entry is 253called "index" and showing it reveals an integer that refers to the 254/dev/hvcs* entry to use to connect to that device. For instance cating the 255index attribute of vty-server adapter 30000004 shows the following:: 256 257 Pow5:/sys/bus/vio/drivers/hvcs/30000004 # cat index 258 2 259 260This index of '2' means that in order to connect to vty-server adapter 26130000004 the user should interact with /dev/hvcs2. 262 263It should be noted that due to the system hotplug I/O capabilities of a 264system the /dev/hvcs* entry that interacts with a particular vty-server 265adapter is not guaranteed to remain the same across system reboots. Look 266in the Q & A section for more on this issue. 267 2686. Disconnection 269================ 270 271As a security feature to prevent the delivery of stale data to an 272unintended target the Power5 system firmware disables the fetching of data 273and discards that data when a connection between a vty-server and a vty has 274been severed. As an example, when a vty-server is immediately disconnected 275from a vty following output of data to the vty the vty adapter may not have 276enough time between when it received the data interrupt and when the 277connection was severed to fetch the data from firmware before the fetch is 278disabled by firmware. 279 280When hvcs is being used to serve consoles this behavior is not a huge issue 281because the adapter stays connected for large amounts of time following 282almost all data writes. When hvcs is being used as a tty conduit to tunnel 283data between two partitions [see Q & A below] this is a huge problem 284because the standard Linux behavior when cat'ing or dd'ing data to a device 285is to open the tty, send the data, and then close the tty. If this driver 286manually terminated vty-server connections on tty close this would close 287the vty-server and vty connection before the target vty has had a chance to 288fetch the data. 289 290Additionally, disconnecting a vty-server and vty only on module removal or 291adapter removal is impractical because other vty-servers in other 292partitions may require the usage of the target vty at any time. 293 294Due to this behavioral restriction disconnection of vty-servers from the 295connected vty is a manual procedure using a write to a sysfs attribute 296outlined below, on the other hand the initial vty-server connection to a 297vty is established automatically by this driver. Manual vty-server 298connection is never required. 299 300In order to terminate the connection between a vty-server and vty the 301"vterm_state" sysfs attribute within each vty-server's sysfs entry is used. 302Reading this attribute reveals the current connection state of the 303vty-server adapter. A zero means that the vty-server is not connected to a 304vty. A one indicates that a connection is active. 305 306Writing a '0' (zero) to the vterm_state attribute will disconnect the VTERM 307connection between the vty-server and target vty ONLY if the vterm_state 308previously read '1'. The write directive is ignored if the vterm_state 309read '0' or if any value other than '0' was written to the vterm_state 310attribute. The following example will show the method used for verifying 311the vty-server connection status and disconnecting a vty-server connection:: 312 313 Pow5:/sys/bus/vio/drivers/hvcs/30000004 # cat vterm_state 314 1 315 316 Pow5:/sys/bus/vio/drivers/hvcs/30000004 # echo 0 > vterm_state 317 318 Pow5:/sys/bus/vio/drivers/hvcs/30000004 # cat vterm_state 319 0 320 321All vty-server connections are automatically terminated when the device is 322hotplug removed and when the module is removed. 323 3247. Configuration 325================ 326 327Each vty-server has a sysfs entry in the /sys/devices/vio directory, which 328is symlinked in several other sysfs tree directories, notably under the 329hvcs driver entry, which looks like the following example:: 330 331 Pow5:/sys/bus/vio/drivers/hvcs # ls 332 . .. 30000003 30000004 rescan 333 334By design, firmware notifies the hvcs driver of vty-server lifetimes and 335partner vty removals but not the addition of partner vtys. Since an HMC 336Super Admin can add partner info dynamically we have provided the hvcs 337driver sysfs directory with the "rescan" update attribute which will query 338firmware and update the partner info for all the vty-servers that this 339driver manages. Writing a '1' to the attribute triggers the update. An 340explicit example follows: 341 342 Pow5:/sys/bus/vio/drivers/hvcs # echo 1 > rescan 343 344Reading the attribute will indicate a state of '1' or '0'. A one indicates 345that an update is in process. A zero indicates that an update has 346completed or was never executed. 347 348Vty-server entries in this directory are a 32 bit partition unique unit 349address that is created by firmware. An example vty-server sysfs entry 350looks like the following:: 351 352 Pow5:/sys/bus/vio/drivers/hvcs/30000004 # ls 353 . current_vty devspec name partner_vtys 354 .. index partner_clcs vterm_state 355 356Each entry is provided, by default with a "name" attribute. Reading the 357"name" attribute will reveal the device type as shown in the following 358example:: 359 360 Pow5:/sys/bus/vio/drivers/hvcs/30000003 # cat name 361 vty-server 362 363Each entry is also provided, by default, with a "devspec" attribute which 364reveals the full device specification when read, as shown in the following 365example:: 366 367 Pow5:/sys/bus/vio/drivers/hvcs/30000004 # cat devspec 368 /vdevice/vty-server@30000004 369 370Each vty-server sysfs dir is provided with two read-only attributes that 371provide lists of easily parsed partner vty data: "partner_vtys" and 372"partner_clcs":: 373 374 Pow5:/sys/bus/vio/drivers/hvcs/30000004 # cat partner_vtys 375 30000000 376 30000001 377 30000002 378 30000000 379 30000000 380 381 Pow5:/sys/bus/vio/drivers/hvcs/30000004 # cat partner_clcs 382 U5112.428.103048A-V3-C0 383 U5112.428.103048A-V3-C2 384 U5112.428.103048A-V3-C3 385 U5112.428.103048A-V4-C0 386 U5112.428.103048A-V5-C0 387 388Reading partner_vtys returns a list of partner vtys. Vty unit address 389numbering is only per-partition-unique so entries will frequently repeat. 390 391Reading partner_clcs returns a list of "converged location codes" which are 392composed of a system serial number followed by "-V*", where the '*' is the 393target partition number, and "-C*", where the '*' is the slot of the 394adapter. The first vty partner corresponds to the first clc item, the 395second vty partner to the second clc item, etc. 396 397A vty-server can only be connected to a single vty at a time. The entry, 398"current_vty" prints the clc of the currently selected partner vty when 399read. 400 401The current_vty can be changed by writing a valid partner clc to the entry 402as in the following example:: 403 404 Pow5:/sys/bus/vio/drivers/hvcs/30000004 # echo U5112.428.10304 405 8A-V4-C0 > current_vty 406 407Changing the current_vty when a vty-server is already connected to a vty 408does not affect the current connection. The change takes effect when the 409currently open connection is freed. 410 411Information on the "vterm_state" attribute was covered earlier on the 412chapter entitled "disconnection". 413 4148. Questions & Answers: 415======================= 416 417Q: What are the security concerns involving hvcs? 418 419A: There are three main security concerns: 420 421 1. The creator of the /dev/hvcs* nodes has the ability to restrict 422 the access of the device entries to certain users or groups. It 423 may be best to create a special hvcs group privilege for providing 424 access to system consoles. 425 426 2. To provide network security when grabbing the console it is 427 suggested that the user connect to the console hosting partition 428 using a secure method, such as SSH or sit at a hardware console. 429 430 3. Make sure to exit the user session when done with a console or 431 the next vty-server connection (which may be from another 432 partition) will experience the previously logged in session. 433 434--------------------------------------------------------------------------- 435 436Q: How do I multiplex a console that I grab through hvcs so that other 437people can see it: 438 439A: You can use "screen" to directly connect to the /dev/hvcs* device and 440setup a session on your machine with the console group privileges. As 441pointed out earlier by default screen doesn't provide the termcap settings 442for most terminal emulators to provide adequate character conversion from 443term type "screen" to others. This means that curses based programs may 444not display properly in screen sessions. 445 446--------------------------------------------------------------------------- 447 448Q: Why are the colors all messed up? 449Q: Why are the control characters acting strange or not working? 450Q: Why is the console output all strange and unintelligible? 451 452A: Please see the preceding section on "Connection" for a discussion of how 453applications can affect the display of character control sequences. 454Additionally, just because you logged into the console using and xterm 455doesn't mean someone else didn't log into the console with the HMC console 456(vt320) before you and leave the session logged in. The best thing to do 457is to export TERM to the terminal type of your terminal emulator when you 458get the console. Additionally make sure to "exit" the console before you 459disconnect from the console. This will ensure that the next user gets 460their own TERM type set when they login. 461 462--------------------------------------------------------------------------- 463 464Q: When I try to CONNECT kermit to an hvcs device I get: 465"Sorry, can't open connection: /dev/hvcs*"What is happening? 466 467A: Some other Power5 console mechanism has a connection to the vty and 468isn't giving it up. You can try to force disconnect the consoles from the 469HMC by right clicking on the partition and then selecting "close terminal". 470Otherwise you have to hunt down the people who have console authority. It 471is possible that you already have the console open using another kermit 472session and just forgot about it. Please review the console options for 473Power5 systems to determine the many ways a system console can be held. 474 475OR 476 477A: Another user may not have a connectivity method currently attached to a 478/dev/hvcs device but the vterm_state may reveal that they still have the 479vty-server connection established. They need to free this using the method 480outlined in the section on "Disconnection" in order for others to connect 481to the target vty. 482 483OR 484 485A: The user profile you are using to execute kermit probably doesn't have 486permissions to use the /dev/hvcs* device. 487 488OR 489 490A: You probably haven't inserted the hvcs.ko module yet but the /dev/hvcs* 491entry still exists (on systems without udev). 492 493OR 494 495A: There is not a corresponding vty-server device that maps to an existing 496/dev/hvcs* entry. 497 498--------------------------------------------------------------------------- 499 500Q: When I try to CONNECT kermit to an hvcs device I get: 501"Sorry, write access to UUCP lockfile directory denied." 502 503A: The /dev/hvcs* entry you have specified doesn't exist where you said it 504does? Maybe you haven't inserted the module (on systems with udev). 505 506--------------------------------------------------------------------------- 507 508Q: If I already have one Linux partition installed can I use hvcs on said 509partition to provide the console for the install of a second Linux 510partition? 511 512A: Yes granted that your are connected to the /dev/hvcs* device using 513kermit or cu or some other program that doesn't provide terminal emulation. 514 515--------------------------------------------------------------------------- 516 517Q: Can I connect to more than one partition's console at a time using this 518driver? 519 520A: Yes. Of course this means that there must be more than one vty-server 521configured for this partition and each must point to a disconnected vty. 522 523--------------------------------------------------------------------------- 524 525Q: Does the hvcs driver support dynamic (hotplug) addition of devices? 526 527A: Yes, if you have dlpar and hotplug enabled for your system and it has 528been built into the kernel the hvcs drivers is configured to dynamically 529handle additions of new devices and removals of unused devices. 530 531--------------------------------------------------------------------------- 532 533Q: For some reason /dev/hvcs* doesn't map to the same vty-server adapter 534after a reboot. What happened? 535 536A: Assignment of vty-server adapters to /dev/hvcs* entries is always done 537in the order that the adapters are exposed. Due to hotplug capabilities of 538this driver assignment of hotplug added vty-servers may be in a different 539order than how they would be exposed on module load. Rebooting or 540reloading the module after dynamic addition may result in the /dev/hvcs* 541and vty-server coupling changing if a vty-server adapter was added in a 542slot between two other vty-server adapters. Refer to the section above 543on how to determine which vty-server goes with which /dev/hvcs* node. 544Hint; look at the sysfs "index" attribute for the vty-server. 545 546--------------------------------------------------------------------------- 547 548Q: Can I use /dev/hvcs* as a conduit to another partition and use a tty 549device on that partition as the other end of the pipe? 550 551A: Yes, on Power5 platforms the hvc_console driver provides a tty interface 552for extra /dev/hvc* devices (where /dev/hvc0 is most likely the console). 553In order to get a tty conduit working between the two partitions the HMC 554Super Admin must create an additional "serial server" for the target 555partition with the HMC gui which will show up as /dev/hvc* when the target 556partition is rebooted. 557 558The HMC Super Admin then creates an additional "serial client" for the 559current partition and points this at the target partition's newly created 560"serial server" adapter (remember the slot). This shows up as an 561additional /dev/hvcs* device. 562 563Now a program on the target system can be configured to read or write to 564/dev/hvc* and another program on the current partition can be configured to 565read or write to /dev/hvcs*. Now you have a tty conduit between two 566partitions. 567 568--------------------------------------------------------------------------- 569 5709. Reporting Bugs: 571================== 572 573The proper channel for reporting bugs is either through the Linux OS 574distribution company that provided your OS or by posting issues to the 575PowerPC development mailing list at: 576 577linuxppc-dev@lists.ozlabs.org 578 579This request is to provide a documented and searchable public exchange 580of the problems and solutions surrounding this driver for the benefit of 581all users. 582